{"id": "train_070000", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "bcfa48344fbf5fec919faee32d98b026f2918d70"}
{"id": "train_070001", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "df3cd0bf14f7255139b07fb89beb35a3a8f4f300"}
{"id": "train_070002", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a8eca372c658a4444b7277a758b4c7f92b70edf3"}
{"id": "train_070003", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "5103645d7533b5b241091e4d420445d28c1f210b"}
{"id": "train_070004", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "6d30d8f0838824038f47b5fac84ed408753dc23d"}
{"id": "train_070005", "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\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "0fe4054cac125ec34cd42806d6a99964d70d7fad"}
{"id": "train_070006", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "166db7376df90c4a3fd3f199858b54b1f235aecf"}
{"id": "train_070007", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "b3fde43b2bec5b7fc1db468cc5128c7333cc121c"}
{"id": "train_070008", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "af800bcc9d57cefc93c725f95d44571b741a9105"}
{"id": "train_070009", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "8971790fdb0324789364a7ad5700c24ad6f1df91"}
{"id": "train_070010", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "a02f60366c1fea2eaf8721b5da52f2e230fa5366"}
{"id": "train_070011", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "ed5013116567f7e5e181b17dcadb411298372be3"}
{"id": "train_070012", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "0524c95c6cdfbdd4c48792072b792d499b2b7956"}
{"id": "train_070013", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "5fd65f8f0ea3f6bdff0b43374c65810e785402e0"}
{"id": "train_070014", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "c3805a3fe8d639ea8b2c377fa0842e8bb0b8a4fa"}
{"id": "train_070015", "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\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e94fcc5c4d24f400ed7f452074479d2267780f84"}
{"id": "train_070016", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "c63612ad248a7a49fe8187cdf818cc8ddf2492fa"}
{"id": "train_070017", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "d27e9048da693de71386431273afe93be0e7a366"}
{"id": "train_070018", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4f3c1a2582eea8af3c79c9c1f963f8e2e6aa2312"}
{"id": "train_070019", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f16d9644b020f03faa36cea6a7363a7cd50df594"}
{"id": "train_070020", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "070148d76120909eb938a017520ea39eacf9ae48"}
{"id": "train_070021", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "46adc5ac26e7e1b2578650719d192969471f5b21"}
{"id": "train_070022", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "b88ad6ae525d1721e46b76230c1c087675af08e5"}
{"id": "train_070023", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "bbc971daf3983d139ed50a1c168043ff5828d0c8"}
{"id": "train_070024", "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\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "8ca9ed843cfa5fcfa607fa11369ad7c29dd32456"}
{"id": "train_070025", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "03a42e1910c3cabf0ef2faa659f7e6652ac42a45"}
{"id": "train_070026", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "1b0c647887cd81527035b48427d2ca08949deb3f"}
{"id": "train_070027", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "7c3a36cf33e36b0400fa2dfe89b44992852345ce"}
{"id": "train_070028", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "026eaab5afdde359f87ed0fc45776c3186192290"}
{"id": "train_070029", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "d3911a5c968bf87aa808883a58e7c41a72e1a4e6"}
{"id": "train_070030", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "c1015ff527e942a0f3535026cacac43bb6ac1942"}
{"id": "train_070031", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["security", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "c63612ad248a7a49fe8187cdf818cc8ddf2492fa"}
{"id": "train_070032", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "7a6cf89ab6df6887d9cc4d91247863e7ff6ad01a"}
{"id": "train_070033", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "3b68b41d5c15b0a13b248c9d8eabbd52383ea6a0"}
{"id": "train_070034", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "aa709c61609590a9cc349cd02bbd72f2908cfbfd"}
{"id": "train_070035", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "edc51437931a5b8ad651f84767ec4d3d36c4c00f"}
{"id": "train_070036", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "7e09c1266f611024ebf089221e3b97e36675eeed"}
{"id": "train_070037", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "366c667e83e68564812c987411a7ee8157d414fe"}
{"id": "train_070038", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "0c3602287f179470fd9b5b00a748a671801f7896"}
{"id": "train_070039", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "8b47ea968a573c9649d3b02145b8bfe5c5de3446"}
{"id": "train_070040", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "6967076578f21ac1c77fe5e8fdaa42f516229654"}
{"id": "train_070041", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "91fbb42b2091872eb47ef6665dc201c4a5b8e801"}
{"id": "train_070042", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "924a9e2a08a65727c35776dbc0508607fcac9a66"}
{"id": "train_070043", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "22cd2b5acd2a6d009be27a1ee5c80ab1d36454c9"}
{"id": "train_070044", "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\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "412cb12f44b0160ee3a237766f23304de63bc3da"}
{"id": "train_070045", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "e252a8b8f9ec096b8da5a33c584c4ebf4351705d"}
{"id": "train_070046", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "2cfd709b5f0b7878635daedd111e50fa63a2bc2a"}
{"id": "train_070047", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "fd318346bcb3b193229cc16f19b50d65386f4c06"}
{"id": "train_070048", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "287a3d1afa6c9d0d627cd0b6d3f346f84819ce7e"}
{"id": "train_070049", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "45c3505a601c417305a8222c482829d4289e1273"}
{"id": "train_070050", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["coding", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "26005f11970a73d3cc939737afa613991f9b7513"}
{"id": "train_070051", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "35d1f44082a29837b8754034980c12361208af6b"}
{"id": "train_070052", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "7f0c7b50fd9135ef9f54e84ab12d27eea10c1ed7"}
{"id": "train_070053", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "b7b8f0372f990b8941209012a03a96776d1139b7"}
{"id": "train_070054", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "240462aaf8fe71b56565a10386e39eea42511e19"}
{"id": "train_070055", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "16ae1594836da5610ec3f2126932c1d6becf850a"}
{"id": "train_070056", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "7adf3c05e4f6fa1950e588955398c9f472fd0cc5"}
{"id": "train_070057", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "e59adea3a1c4ea8f9e75c9fcd7ac227e305700e5"}
{"id": "train_070058", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "e46449e4a4df9b7c405455f95c6e1afa4b1461de"}
{"id": "train_070059", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a626003ad6efee3efc2b53034c134b50390c8fae"}
{"id": "train_070060", "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\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "500e4d9ce3d8edc5c737859ab0892c5371c95ddf"}
{"id": "train_070061", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "0fc1a5fbe49076875ea39595fa6e0223578544a0"}
{"id": "train_070062", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "effc0792121a524fa5fc7bb5e51ef65505263e2d"}
{"id": "train_070063", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "39e5fd9c402d9efc514e93373dea02e6dc1e077e"}
{"id": "train_070064", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "686fba71ceca27e364d3372d6ab536e5846172a4"}
{"id": "train_070065", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "72d0a39559afcb979863815d90f50efeb23caa1a"}
{"id": "train_070066", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "443b5b1f9abdae35552c215ac78a892090f00be8"}
{"id": "train_070067", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "ac6453bd6eac93869cdcf7b4d09cbe8082c168c1"}
{"id": "train_070068", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "11d5ae5353e5a7603021ae00cffc8bf0d9847fa1"}
{"id": "train_070069", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "9203009c6691a89cc380f60cdfdbb4d8b5ad2577"}
{"id": "train_070070", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "7608e20f37f5d3c1b1e078bff8a76defa548f5aa"}
{"id": "train_070071", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "5354221c80eed7a045d06efc6768d242244795a4"}
{"id": "train_070072", "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\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "06a2cc864656b3a48a37881caf3b41750dba19fa"}
{"id": "train_070073", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "92b5e6b54d7a74ffb0073192a0f63b1b0c266cc7"}
{"id": "train_070074", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "50aa57d1287460ff41bec4ed9ad04252018ca431"}
{"id": "train_070075", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "400ed9801dd4fdf8bfb90082a6cf760027cb1421"}
{"id": "train_070076", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "276f875f15a86f114c31830292155314fe3e494d"}
{"id": "train_070077", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "b5ae4113bcc864dc808190261c19fdc046fa4ff0"}
{"id": "train_070078", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "ee15ffd4255a04a642601c3731bae4fa64b00019"}
{"id": "train_070079", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "1fe9837b2777b42c2b7ebc67ee2f8958b909c38a"}
{"id": "train_070080", "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\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "6602cea9d988a5979b957e478d40c81f408cbb78"}
{"id": "train_070081", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "89709c1afb1d85db53f088dd5361634adf11f42f"}
{"id": "train_070082", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "051df7b5a82096e64b216158f0220a48df2afba4"}
{"id": "train_070083", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "41515788fa36682599525fb4a59d14c070d2a893"}
{"id": "train_070084", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a42c5c3db261fa1065ce27ea844d090524e45a4c"}
{"id": "train_070085", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "9e19ba83d05d2d5b585435f0613c596049fc284f"}
{"id": "train_070086", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "ff7e3d7bf7c82900756453daa2b3441ed2dda585"}
{"id": "train_070087", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "67dba8ffd7aa9973b9c44ad2a6cd67cf1c800cab"}
{"id": "train_070088", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a8e968982663fcbf14699b954df55103d52c3f69"}
{"id": "train_070089", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "d238340776354878be7e7bfe4f44f08b63ea80a4"}
{"id": "train_070090", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "94e472a1d46d1574a72f2b9ed13979803662e167"}
{"id": "train_070091", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "5060ab567909a0b008bd2751decf11b954c1e704"}
{"id": "train_070092", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "50616df38e5e36c02dbb4aa90e285f5691d31a52"}
{"id": "train_070093", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "f785e2a1f58bff737d815a1380f492c651a6c1c3"}
{"id": "train_070094", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "5a933a8831506f9f7e8aa756b369cde57e368915"}
{"id": "train_070095", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "176569cf26c4062e5a824cbacda81da2d4fa6eae"}
{"id": "train_070096", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "c680235903a664e96495f36a169bed14b2cf9e71"}
{"id": "train_070097", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "76f8b6577a36eca868c2edbcf145b3533b5cca61"}
{"id": "train_070098", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "eb287ab03c21ab30398607e864e13573b08f1522"}
{"id": "train_070099", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "8b705a0d5e913bf9bea8dd670af4ae9567f7a1a6"}
{"id": "train_070100", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "a1e8e5d54d1913748c6c366188a3f7f1cce3c61f"}
{"id": "train_070101", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "1d993443d75a272b4719925bba480c61a199a576"}
{"id": "train_070102", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "af19f25e93da6ad38276c6d20e1e6d7a164b84cd"}
{"id": "train_070103", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "1bc2ca8386bd300096bba10f11a1f480ed24dfc1"}
{"id": "train_070104", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "cd022c0a4d5d3e16ab08cc4bd6071e4c317e8de4"}
{"id": "train_070105", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "6a9bfa92458d830be232f40522c46cf2b0cf6698"}
{"id": "train_070106", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "e6049562c126c309aa15620b72a050c609589c6f"}
{"id": "train_070107", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "fe408ab6b3d1a97730dd41cb26aaf4848f47c505"}
{"id": "train_070108", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "05ed4f5e14ee60b079f57de628005d0b42037fdd"}
{"id": "train_070109", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "80bf3a9a0d4e734c18ac31f57d1a52f552637077"}
{"id": "train_070110", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "7ebe83fa77e4ba1b4469232923fb26685a886a61"}
{"id": "train_070111", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "a4b9c79e0bf4bf4a4538e75b99ba90fffec854f3"}
{"id": "train_070112", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "1ef0506e100603eca5ccf97caed621512783a8e5"}
{"id": "train_070113", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "2b3fc26b3428845c6aead6b3cfe059f079caadd6"}
{"id": "train_070114", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a4c97191e0f76931c048e97b33a02be1938e8049"}
{"id": "train_070115", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "f75c69849d56573e691e7130371f5a415df02533"}
{"id": "train_070116", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "28c219244247278d4aa622280d4a3e77588ca4e9"}
{"id": "train_070117", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["evaluation", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "b917effc9cc21f0a09f5b9c24d91e47b71f98f3f"}
{"id": "train_070118", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "e0f8d254e8671da347d69a87208532ce9d91236f"}
{"id": "train_070119", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "1feda04145ea2a14cd431293f6b8f333c6a89b78"}
{"id": "train_070120", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "3a0bf52d3c1e718f3e19855374fb57dba850e633"}
{"id": "train_070121", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "717dab252f8cae30f68cd9fa3f6f4b50be66b87d"}
{"id": "train_070122", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "54f2bdccdf3e8188a3acef6fdd28fa636a967bf3"}
{"id": "train_070123", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["governance", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "90bd26758ee1b66e43eaedb5159161a36bfcb635"}
{"id": "train_070124", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "2ebac9e136a595748fc8c8852a941f0f65c694f8"}
{"id": "train_070125", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "646735f7dc0db4ac01ea3cc12b9c2b25a9a1b4f1"}
{"id": "train_070126", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "8316fa8633082ec7dbbf8a299fb6506c6d71d419"}
{"id": "train_070127", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "3f17779fc50cf8a460906852f72cf52d539e6019"}
{"id": "train_070128", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "9edffdab2dfe17bdea443c6b673ad105cb23ce6f"}
{"id": "train_070129", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "260076bd3c915a6439e17040820dc5a3445a530e"}
{"id": "train_070130", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "0b283dd9352ee67cfbaa23b4cca46eee510708dc"}
{"id": "train_070131", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "7ab0a2de81fdd6363f62655faea67fbfe03583d1"}
{"id": "train_070132", "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\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "c96ba243a8852f67fef5d2d6d74d8d9b6d025f20"}
{"id": "train_070133", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "99d13d10444739252a459b05fb4e5286a1530a51"}
{"id": "train_070134", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "52c5f2273c5c6b3c29857afce1e964325f6ec6b2"}
{"id": "train_070135", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "c1015ff527e942a0f3535026cacac43bb6ac1942"}
{"id": "train_070136", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "35d9baa629c894f66c7965c5efca936a75317d43"}
{"id": "train_070137", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "cc5fbb07b356de6661767ed407421a1437cc2bd4"}
{"id": "train_070138", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "1ac3f35a57787d323c1e3df25b48f46d1f3686c5"}
{"id": "train_070139", "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\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "44837270b443e2017473c0e275d405f10c3750f0"}
{"id": "train_070140", "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\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "5ad98cacf7678b3ebdbd18e8c55ea6a5f211e550"}
{"id": "train_070141", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "07947f846c569db5250fe5bf129cabdbf09e3f07"}
{"id": "train_070142", "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\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "d7c57de796694f28b89ca762aeedb9cd321759c7"}
{"id": "train_070143", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "896a585baf3ab86e1b0c86185485ebdc542e2dd6"}
{"id": "train_070144", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "7e411553f2cb167f27443687c3b141990dabc04d"}
{"id": "train_070145", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "b053c0473250165086d91b3394b2164a45ad8575"}
{"id": "train_070146", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "635e3d6f99728ffc04d43ba5ab9f4752536979a1"}
{"id": "train_070147", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "5d7f06c6b34e5ec4943b640ff157fc6ce99fc77f"}
{"id": "train_070148", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "b3e37a7fd8f2840a691664c3ee3283e5be52926e"}
{"id": "train_070149", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "cc192838215640c59e594f1e3d5928ff710e5002"}
{"id": "train_070150", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "bdd778786ee22fe0abb659a4c062ef9091d1a1d3"}
{"id": "train_070151", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "8fc8d6616db2578eccb103b58a88af1c0c035e1b"}
{"id": "train_070152", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "d462be2b439fb8eeb9853d2c7d553dd701911351"}
{"id": "train_070153", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "4f55f4d876656593445686ecca258edab0a08817"}
{"id": "train_070154", "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\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "20f656c3d51fa71d36b1c1078c1728917572d20a"}
{"id": "train_070155", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["agent", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "705daa25c9528b31a93e8d3750d10490d16ac1f3"}
{"id": "train_070156", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "1daed4e610284651e4b15de2231122e262bba60d"}
{"id": "train_070157", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "72c7bb5efe3b3d4f514ab57815fa81939ac03501"}
{"id": "train_070158", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "4a5bd948c3e3ca5f7e7af9d0b0f1129f30f3dbff"}
{"id": "train_070159", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "cbbc23f1d89d0f0387729a4f6ea3031714dc6f40"}
{"id": "train_070160", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "4c8d599e1b7de9334e858b4ac12068861a9836e7"}
{"id": "train_070161", "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\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "d996264d983392bfab72a2ab213724f292185c83"}
{"id": "train_070162", "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\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "d839bbaffbb450988087e2f3a24ced7f49345a27"}
{"id": "train_070163", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "a0deeb729c4589546c2b08c989917d5fcd979391"}
{"id": "train_070164", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "705b969e1931a8c25d16291fbe45cd96026c3f08"}
{"id": "train_070165", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4dd935adcfc25b30da7283b04df12dc1f7e8c407"}
{"id": "train_070166", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "c69d709438c534ce9dabec030f6031e2f37c519f"}
{"id": "train_070167", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "558c10bccdcf2f4bbd644b4ce78154d2334dfd50"}
{"id": "train_070168", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "2bd66369e420db151d0ff41493dfeb0e82cc7e03"}
{"id": "train_070169", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "4169c8f23535e35bbd9a5960a51abc0be8493aab"}
{"id": "train_070170", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "8ca683f8d11247b9f39e7670c6549dabd12a7130"}
{"id": "train_070171", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "93d21933b237640a504763aa08c67d3839dbe3ed"}
{"id": "train_070172", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "052e30ed22a5decff3cec73cd8502265a09a8dec"}
{"id": "train_070173", "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\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "5e93fbba6e575fc5894214c03785f59d424227a8"}
{"id": "train_070174", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "fe3dd7119fc96afab48e357475d48cd3c7931d07"}
{"id": "train_070175", "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\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "6ca67b94549c95987e6a4ea3466691f104b9f773"}
{"id": "train_070176", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "d783ab821be2019d5a7449dbfeb35245b4adaa04"}
{"id": "train_070177", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "740e12552ed276d9501098bc82e9c8c302a992a3"}
{"id": "train_070178", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "246f3378872c930b7f08b77db47ad33b837b6857"}
{"id": "train_070179", "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\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "fb41125ff1c750bcdaac8d542a050d35ef384286"}
{"id": "train_070180", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "986cb1fdecfdaad6d2b45850e0e7c431c5f9ded5"}
{"id": "train_070181", "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\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "bac31559ebd094d9799770a21eeaaa0dd93e7245"}
{"id": "train_070182", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "c3c326a91418d3ff95bc91caea0bb5430e8bfc1d"}
{"id": "train_070183", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["governance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "6500953aaf26a9f82f2e6909ded1ae85da83a3ba"}
{"id": "train_070184", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "550c4d7176fbb91170501aa483075278b3422ab9"}
{"id": "train_070185", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "12f1c398bff392eb317431da737910b34b35a734"}
{"id": "train_070186", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "97cff8d8601d404e37d9ec81346be06b259fbc28"}
{"id": "train_070187", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "828950b478927cb1238c7a0448fe7570fae1332d"}
{"id": "train_070188", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "6b03b858c0fe6258be4ca58754ebed406d674c89"}
{"id": "train_070189", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "70f5c75a9acded627fccc423136ba89a893fbcf8"}
{"id": "train_070190", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "6ae5e368aa53b64dc4c0f48625bbaae2702ce3fd"}
{"id": "train_070191", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "fc74af4c92d2f248306a2ae1b951fe2104ac1498"}
{"id": "train_070192", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "2c21b3902b1fddb2ab698fae8e14b585cc989630"}
{"id": "train_070193", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "1cbae78cb4ccd7077d97b74ee7bbfdc491f7babe"}
{"id": "train_070194", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f91c1a1e1ce0f692d7a971d8f146aa097fd76cf8"}
{"id": "train_070195", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "d6f2e459424d7d4335d5336ad44250bf5301c52d"}
{"id": "train_070196", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e606cca094644f49ca7767542f30303147a62c69"}
{"id": "train_070197", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "3a2770b70f682e40a455fb5d21cb05dad05e0d6a"}
{"id": "train_070198", "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\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "5fdbd84f5662cf4f43722caf9cc782cec7eb3c35"}
{"id": "train_070199", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "a0dca6dc7bf87e85762d77a2aaae252f7fae9ae6"}
{"id": "train_070200", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "6b393496ca8fec8334ba2626e0b114c5f1741adf"}
{"id": "train_070201", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "d24c60737921f63b931ec208231f528c3d75e595"}
{"id": "train_070202", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "a2773dea1b42186e6b70778ca93516f563516eef"}
{"id": "train_070203", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "f9f8909c628b6fdec2fb3ea155bbba8b32e52808"}
{"id": "train_070204", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "ac81d0ddda075c6662bad332dcfee5d3917229fe"}
{"id": "train_070205", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "82b03c12f982747cbc1e2f74ead3fe0b9ffc4e45"}
{"id": "train_070206", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "638a76f9c4d2c1b848109f389ab21e3465eb8389"}
{"id": "train_070207", "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\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "751d5b9e2965825ac565726e25709baa9df614d6"}
{"id": "train_070208", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "302ce8822af15052766544a1983cf027d68db59b"}
{"id": "train_070209", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "c6a018fc6a6e0629efc87652337c153cbba0531c"}
{"id": "train_070210", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a097d7d42ee89ce25667fca7b37a5b7da0a30a78"}
{"id": "train_070211", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "28da167b0f912c90c4d5c41885c608ec43a60f16"}
{"id": "train_070212", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "8d8bc925cf8204ea011da25b605c0137987a4aac"}
{"id": "train_070213", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "7f488ba2ad69da98c8b6fffb7f32e4590509f1ca"}
{"id": "train_070214", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "fb527e5a729a7f588d54c3276b7d1b8d50404c2f"}
{"id": "train_070215", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "8bbf74f2cc32fec9c897b2f4d9eab45b0abf3053"}
{"id": "train_070216", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "50db4a4080827dbbaa4ee38b8432d5ce3141f6e7"}
{"id": "train_070217", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "2da4fc709ca7001eb874c74f449bdd0a56b07bcb"}
{"id": "train_070218", "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\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "acd25e5082e0e00e09d44e99c7f889c9b1419519"}
{"id": "train_070219", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "12fc873e4196bab4c560a95fd313099320a31ef7"}
{"id": "train_070220", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b17ebbeb1e8f571cdc634d3e951a2329b5e3f0cb"}
{"id": "train_070221", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "c2102c2aeea58f42f30346234f790eb9679ffeed"}
{"id": "train_070222", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "1e4c37dfc7eb1fbd2b2903822334157fa7dd6551"}
{"id": "train_070223", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f8eb40a2a05c85395626382ceafa5dd3791861dd"}
{"id": "train_070224", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "9dbda26c9d46c52924257c791d331ad0dc4ebb5f"}
{"id": "train_070225", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "2ad6dbfd7dc423b907f65ae0e99243116e68b563"}
{"id": "train_070226", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "513cb9fdd399b5d83f247486f45f97b2e1bf3701"}
{"id": "train_070227", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "01c3ce1e07faa4fb9c3d1603ec4edc2f06371b6f"}
{"id": "train_070228", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "311949814daaa2e4a0d464b235984c06d30246b6"}
{"id": "train_070229", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "5008de4f627e6f0eb81acb31985ca9cc03652cc8"}
{"id": "train_070230", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "1f0853ec5796e20dd49c336bba59588eb4bdac02"}
{"id": "train_070231", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "0876d6937faa40f930730942357baa5ce0ffc8f8"}
{"id": "train_070232", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "de91afe55414cf6a15d2ae5a9f97b6f1580e5c8d"}
{"id": "train_070233", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "4d9899f7dfe82c43ae802740f1e49451cf47e972"}
{"id": "train_070234", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "6be3475790b646c71d3032d613266c56289d9aca"}
{"id": "train_070235", "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\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "a232cb3759bedbf84c59f0cfdd692dd6c351b206"}
{"id": "train_070236", "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\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "c854dcaa20edd575dbf70932917de978acc3f410"}
{"id": "train_070237", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "259eb9185b57e0fa13d7854f011079367f2528a2"}
{"id": "train_070238", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "0e855dbcb4546d3176e59359547f636a89601772"}
{"id": "train_070239", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "fcb9a78572776723708dd86daa13cc676edf22cb"}
{"id": "train_070240", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4d0491058305713b962b7c698c443f883a7f789f"}
{"id": "train_070241", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4bde4bffc0529330211e0f636cff4c6e84e80929"}
{"id": "train_070242", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f4e053186f9703c680f9b4aaf43a07d47982a53c"}
{"id": "train_070243", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "08fd6adb2530ece689e2e3b501aec81239ed10c8"}
{"id": "train_070244", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "6e24dee98e1af62dc9ee8b1f8bf41a078a889661"}
{"id": "train_070245", "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\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "d8f37a5e60f037147c4bf6f5e35e25f10a8c1f35"}
{"id": "train_070246", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "e9d59dab40cd8980b36bfe24f1a364f37a499d65"}
{"id": "train_070247", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "234058a11fd0fdb1a307bc8c0f3995c8fa775b91"}
{"id": "train_070248", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "b915f8beadb6a92542ff01af4cbf9e39d7d77851"}
{"id": "train_070249", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "cc51b8bc1539e4a7a9668b824cbe452b5eda378d"}
{"id": "train_070250", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "d5c8870c80f696e1bf399c4eb32ee63df7e403f8"}
{"id": "train_070251", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "3ec64cdeb02192f65b935432ed422878d36ae9ac"}
{"id": "train_070252", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "3e50cd8eb0625b74b51add9ea221ccdae74fbfc4"}
{"id": "train_070253", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "516da9d4dfa3767f9e650921a97a31d465e3a94a"}
{"id": "train_070254", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "4b5dbf01714a9bc4304ef8891704c3b432d619c3"}
{"id": "train_070255", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "697eaf47cea15e6783a980363e82208c91db4969"}
{"id": "train_070256", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "ef11e0c22794c8d9b31a0c0d20e6cc048fcfcee6"}
{"id": "train_070257", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "65ddde9ac07963ca551be9a5c6b9ea906bcedf6e"}
{"id": "train_070258", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "1dbb3d4a0ffe8f2e9ecb52a71f7e9b18725280bc"}
{"id": "train_070259", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "0511033023441f519ba6545d78de82e5fd688b20"}
{"id": "train_070260", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "73c2813380e74e014e98768a015f07cfad56f367"}
{"id": "train_070261", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "a665c7853ea1cb16bf5785721d9723e81341c9c3"}
{"id": "train_070262", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "8849f292a0951269a193ff64c2fdff804f73e32c"}
{"id": "train_070263", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "f2782ccf756f48f3603a7f143727923ade75777f"}
{"id": "train_070264", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "7bcedf4228cd0a5b721b60d8da8721976c831d16"}
{"id": "train_070265", "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\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "44d49c373d545450c6e0ea43878b05aaff376fc6"}
{"id": "train_070266", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["research", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "23743843cb24c66e28a903224bda8091ae870231"}
{"id": "train_070267", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "568b3f5885667b29848d17a90944f453c63c74b2"}
{"id": "train_070268", "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\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "64ffa0aacaacb551e7d303cd9bce4a478f278d23"}
{"id": "train_070269", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "d6b2d5cdcfd44c0adcded8916c59389254b4f409"}
{"id": "train_070270", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "aafcd12689ff6f0b887bf3970f24e4be7007488c"}
{"id": "train_070271", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "fcb5f489e544ddb841825419509960da8cf327ce"}
{"id": "train_070272", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "f4622bf8fb7c7f5452a2bca2c8d61bf787bf45bb"}
{"id": "train_070273", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["performance", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "1cf339998bfd226663070822199d2d5452e60eab"}
{"id": "train_070274", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "d8577b97b3f05ad0667926738645502dd40d049f"}
{"id": "train_070275", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "6c88bcbb846ba50779200231ae98f1b41d6ff450"}
{"id": "train_070276", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c9996a5e57c59ab7a19e9bfc20ab9198d9f6fb74"}
{"id": "train_070277", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "399067b0189f378ef805b0364255a53c18e40b0d"}
{"id": "train_070278", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "2e2755c2b1958b8bb6fb317b73b4de5017ae3a11"}
{"id": "train_070279", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "1b1dc26e6603f665e7b436a8bfdc11b2a00542d3"}
{"id": "train_070280", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "70e99adf582b30a8bbe0f36e7291270add81ab7e"}
{"id": "train_070281", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "fb19854d1e4cbd11a5b9c8c329c29d6f28f56c13"}
{"id": "train_070282", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "446376d0428adaa62ec60f21f295e9e9c66d06bb"}
{"id": "train_070283", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "1814a4724b63fad159dbcba5c14d14c7b5f2567e"}
{"id": "train_070284", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "abc5e3b413f45018d8e29dc476c2112727b2382a"}
{"id": "train_070285", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "37dc933fc7c8c44b48f10535ea6b2a336e52fa02"}
{"id": "train_070286", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "a4c8a788d759506d0c3b74d04f6b599576e54b3c"}
{"id": "train_070287", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "5bb3aa0b193eb1a383f6166b6464304456d5dd7e"}
{"id": "train_070288", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "06aa06577dc7d583de699fd821ee3ca73794b1bd"}
{"id": "train_070289", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "79f8d0cc59e4f8921a86ae6f5ea1e7838f36a83a"}
{"id": "train_070290", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "7668183fef35ed0c2838d86c8c38b59225100651"}
{"id": "train_070291", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "7a0a2936414baa22412c6c1d72dd306538d3a192"}
{"id": "train_070292", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "7588740cad867d7fd3f1faf2a510486fd782aa05"}
{"id": "train_070293", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["governance", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "46eaf85ef898306bc5b01c7be8f09b9fad95b89c"}
{"id": "train_070294", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "42de4b85ce8e3a87489a0acb049494df1ab9be2c"}
{"id": "train_070295", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "49d5f095775a18a39ac69e25258887627dd11941"}
{"id": "train_070296", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "8664758a35521975ad7de481bc74024dd5aa166c"}
{"id": "train_070297", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "fa950cb7ab789df9bc266f64bdf9e349eea393ef"}
{"id": "train_070298", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "0344448b8276cb63c91d6f06c16e78bc6442b2ee"}
{"id": "train_070299", "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\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "a07f1a2f16fd622aed357a8b8e20a81b66f98b4b"}
{"id": "train_070300", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "3e7e741b7cc7e5d12d11cf6e0deed2bcc2a74115"}
{"id": "train_070301", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "2fc505308aec4467dc0461b8ef6a266b7f59fc2c"}
{"id": "train_070302", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "e7d6a3e70f195ac9fd6e60be2f75aa0dadcb8cb1"}
{"id": "train_070303", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "3253cf2eb112429f3fc76c39d56c966cbbd32d04"}
{"id": "train_070304", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "cc6df557a29efc50aedfb16732b1b4a57e069973"}
{"id": "train_070305", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "9efd1a454f7404099acfec4dcd088a8b4bd0b328"}
{"id": "train_070306", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["performance", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "9d544d219fdb3d761c49e545dd32410b4edce654"}
{"id": "train_070307", "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\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "1d8fb48db7541283ba1b74a404d13da60585e1c0"}
{"id": "train_070308", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "26926dbed6352d1f4fa2afb69c1d359956476504"}
{"id": "train_070309", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "5779f28722042072506c7d9dfbf11b76c917a089"}
{"id": "train_070310", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "c8dd13cc3468b27b737a9f38c07348cf1abc995c"}
{"id": "train_070311", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "532b024a7d6aadcf4d2255c65952575a7fea18f5"}
{"id": "train_070312", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "d9038fc0eea46a44e565f6a8ce9d98e5f03a8097"}
{"id": "train_070313", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "54f578a4daa898aa4bab5615d33ca23bd84e27c0"}
{"id": "train_070314", "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\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "79e52883972f69aa1d3df7ee4ae46a4e13c527a7"}
{"id": "train_070315", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "0dbdfb7219bda6e40a977853b3739d6149a34a8b"}
{"id": "train_070316", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "c095edccfd191e392c70ec4fe01f75825f2e594c"}
{"id": "train_070317", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "f5ad4b86dd5acb96763abdc4eb3ac1d087eb1051"}
{"id": "train_070318", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "7f6cb8f43ea8562573f4e77571db7ada16e64dd9"}
{"id": "train_070319", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "64497fd4d78e661630a4c64f90a343927af02c8a"}
{"id": "train_070320", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "4e490d34ef737b8ac4e4be809c514c4d517615a8"}
{"id": "train_070321", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["governance", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "db72f324dcd9fc0f57bb645b8366946cf0845d38"}
{"id": "train_070322", "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\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "9d69e434091dda0a4b27efbea148dc2fad5a039d"}
{"id": "train_070323", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "c5d33d84a1fa839df7fa7936081baf0927eda7c3"}
{"id": "train_070324", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "8b0d1d56e58a5f7c30d923aca16675c2aa029003"}
{"id": "train_070325", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "3124eb984ba349ad49719d6beea7f4874befe4d1"}
{"id": "train_070326", "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\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "763cde722ac163fa764b888f0a0b360345700d70"}
{"id": "train_070327", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "775995559d54a596609aa0440fafac171ad4fce5"}
{"id": "train_070328", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "b9228577d212b9f16e299f490e6e889b60dafbb8"}
{"id": "train_070329", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "7c3a36cf33e36b0400fa2dfe89b44992852345ce"}
{"id": "train_070330", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "7130dc4d7d51c9bad96c4b15a84c19a2f0337eb6"}
{"id": "train_070331", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "db9329294ac1d5e2be9ef213d00176830f8c6ea3"}
{"id": "train_070332", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "08120dedd6857a53b5792af228a87c60b5ea7a57"}
{"id": "train_070333", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e9cc55aed05a6fb94b12da27e32ae1faf95afd30"}
{"id": "train_070334", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "7f5a1e9a566e26a90635f8f8ff247c902a4ea957"}
{"id": "train_070335", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["performance", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "61a126decf60763e8d5c7208cade53b9ca6f0268"}
{"id": "train_070336", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "2be687ebc56c16b04e417f5bcd5d7869934f3dbb"}
{"id": "train_070337", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "1a958eeab662b518fc625a7e96c0b60a5e738520"}
{"id": "train_070338", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "3a29c482b2f158159291abe97eeeb2b815c8a832"}
{"id": "train_070339", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "5ed50501dc4eccb1180dd999b2804d8227d0bfa2"}
{"id": "train_070340", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "3a2712b455b262fdaad663f88965fb2016df1d60"}
{"id": "train_070341", "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\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "47034eca184aa87e0b7f4f02b4b12ffdf5797d98"}
{"id": "train_070342", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "c4430fdcfb9b72a6384adaf033fe7364efca5d21"}
{"id": "train_070343", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "438aac7594eb4d0f3c778d76e3f839fff8e46844"}
{"id": "train_070344", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "dc5d9d9a0e89a7f3e7891928275ef1d3cd212502"}
{"id": "train_070345", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "0730147dff0173be577b846d7568a06040045f92"}
{"id": "train_070346", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "55da5c8331ff707b3112dd117a4d3bd0d95a52f3"}
{"id": "train_070347", "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\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "a261ed9af71aabaa217ff894352b1bf9f0815df5"}
{"id": "train_070348", "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\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "c96ba243a8852f67fef5d2d6d74d8d9b6d025f20"}
{"id": "train_070349", "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\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "d996264d983392bfab72a2ab213724f292185c83"}
{"id": "train_070350", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "3d79b19abcce8441c1ce427b83e7fece379a0cd9"}
{"id": "train_070351", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "76731075d4eb4e5b69d996eb6f85f8787280b96e"}
{"id": "train_070352", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "0b0322fc8ccb919922117b56f26ba832afb4e8fc"}
{"id": "train_070353", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "690989dfefb01ac012542e86bc3d5617a72b5480"}
{"id": "train_070354", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "d4c234d6015c538566ee98c7e315775eff4e3446"}
{"id": "train_070355", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "79aac9a02b8ad505f8d8c3fe5647c43c4f49a91b"}
{"id": "train_070356", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "446376d0428adaa62ec60f21f295e9e9c66d06bb"}
{"id": "train_070357", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "d1a4c1f2f2c72b8c231904712520626843f12665"}
{"id": "train_070358", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a6a5630c76c917bd7248a4c92a5b11b0396e4459"}
{"id": "train_070359", "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\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "5ad98cacf7678b3ebdbd18e8c55ea6a5f211e550"}
{"id": "train_070360", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "1251fdc3a169e0126017c07a5a5b6a96bca49034"}
{"id": "train_070361", "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\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "00ca8b239c567cba5a4a84e9d43e894450b5ecf5"}
{"id": "train_070362", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "e0f10be60e6045a847e55280ab23526f811c3726"}
{"id": "train_070363", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "79027e8d556a3d5521567aef4f494c39f95d16b2"}
{"id": "train_070364", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e9b39cd0c7dedaecd8960273646747197d086c5d"}
{"id": "train_070365", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "4b3274a4f36d5f11bb2371ba548bb4fef25f0828"}
{"id": "train_070366", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "931d041ce852320e0cb8d0c0a13746d7803f27e9"}
{"id": "train_070367", "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\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "4f086522f1d7c9c9ec1bd9270861c8426a31b3ca"}
{"id": "train_070368", "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\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "586e6f92f9ef2e1c7f5ee7ac9bfcfcdd80dca82a"}
{"id": "train_070369", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "da57634387dc3b57e38cd1383fb8ebc1a4396dfa"}
{"id": "train_070370", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "3b214d8c3d0fab7a440ee93fbf8291853135e7d0"}
{"id": "train_070371", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["security", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a1e03d118111cc2a72b39e718a39fa4e1b8c5819"}
{"id": "train_070372", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "afec30f18a9c74767f1860c531172b19f2786134"}
{"id": "train_070373", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "20973a2f759110629a7c3c97ae2425bd42529e11"}
{"id": "train_070374", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["coding", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "3875ac564e48832891544e3efe68a2acfbd14cdf"}
{"id": "train_070375", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "91f59eb0252b388715e1247d23a9acab4fa3d89c"}
{"id": "train_070376", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["governance", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a536d6c65dbab3de94c85d4a98361a4af6fb1cb7"}
{"id": "train_070377", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "fb645a70c140e21e5bb772aedaf8db715652ff20"}
{"id": "train_070378", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "b280ac9241e95a1b3428a527f22138677b720754"}
{"id": "train_070379", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "30935cfc7ce847c791927a578f1966a4637689c1"}
{"id": "train_070380", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "ac6a1e826e01edbba9bc1d81d2caba71abcc2f54"}
{"id": "train_070381", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "9965bd68ce5e8450e4c9ee2183322c8b2a52d8bb"}
{"id": "train_070382", "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\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "c85f6326d644e1cd631a5c1a6e180ad160a24729"}
{"id": "train_070383", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "812865355dd7f880079da1d8a1d012d6ccfd4ee1"}
{"id": "train_070384", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "672779d1c363f5a05e68211522d4143faeedb3ba"}
{"id": "train_070385", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "8907c41f67ec18dd49292ab809eddf0bb0739887"}
{"id": "train_070386", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "b269a5c3efc95e97ccb22a1d81bfc44ba426da52"}
{"id": "train_070387", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "9fb043d814be4fb85c00ece288cdc9f0abc1fbe5"}
{"id": "train_070388", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e6badc166a8c31d68aa9aae2bda6e39f1f71a1be"}
{"id": "train_070389", "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\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "964c2c3bc25a751327c15380010a1303a7834ad5"}
{"id": "train_070390", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "4c7e01f21b1f340a4616b7ff27735c25de82fa6d"}
{"id": "train_070391", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "a6c27cf9786fb752f3bc674682a1ad37fb9eeffa"}
{"id": "train_070392", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "436f79c2ff7d360e184bf4ceac5decf81d4f51d0"}
{"id": "train_070393", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "2ba02266219edfe39b4160b5d789701f84811c35"}
{"id": "train_070394", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e1a2e3b20fba4d6d2490067283360efbacf1736f"}
{"id": "train_070395", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "e9b39cd0c7dedaecd8960273646747197d086c5d"}
{"id": "train_070396", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "9526fc5408ed2389fc19f5fe0f4260bbfa335500"}
{"id": "train_070397", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "426aea6f117d5ce95b830ba99c54e3769d6b0c87"}
{"id": "train_070398", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "b535befffcb30c5bf4a071acf897ed533a3fcd60"}
{"id": "train_070399", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "a368f13ae2a8d36d135c6656eb8ae3e3a9ba58a3"}
{"id": "train_070400", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "57e407347ed2cc267630c67db7b99ba116126d54"}
{"id": "train_070401", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "9326e1c8e4c8784a75badf77ecccf98c62582f97"}
{"id": "train_070402", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "d09b08a2ce53401aacb0ec34bbe38f8b257baf8a"}
{"id": "train_070403", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "b3f99047dfd90a0d5db4ede4910e241d96dda28e"}
{"id": "train_070404", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "381b48a3960591eb9e555734c4f4b40aefa646e1"}
{"id": "train_070405", "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\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a5f05ab1f421cba1d5fa60b07bd2901180763af7"}
{"id": "train_070406", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "64a655ffbf298d843456b63f7bf1c3216d057de9"}
{"id": "train_070407", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "296a50a57650c5fb4ef3c82abdc814fa01208e5d"}
{"id": "train_070408", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "269bcf764180e550d2894d87fd208d2e0d4be3ed"}
{"id": "train_070409", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "5d5c4f56528f591947c93289e967ae7288e54e05"}
{"id": "train_070410", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "58787be06985ee26ed65f13e1f9303e03f5a5c34"}
{"id": "train_070411", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "b3fde43b2bec5b7fc1db468cc5128c7333cc121c"}
{"id": "train_070412", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "2e1e2b97cd592ce5b3d41d8bc39b1b0e9cdb8d31"}
{"id": "train_070413", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "ca712ade3d3ef418ddfeaf398699cb9a4571c358"}
{"id": "train_070414", "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\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "bc88129c6909af849770a79492889507c3bde707"}
{"id": "train_070415", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "634901acd3fd98a281011d6ad1857cd23012822b"}
{"id": "train_070416", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "05c6ee49da93b2175ebe995fa135d43bb1a84fc8"}
{"id": "train_070417", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "666c14144b7f874bb2449e0a802563f0282e02a7"}
{"id": "train_070418", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "17e6f90bd1e38b9a6bff00b22a6a518e0b13d03f"}
{"id": "train_070419", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f818f8dcdd682813b0bf683c4d44e957122d97db"}
{"id": "train_070420", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "ed528c9bce17f3b408b4552b246d2efd5a112770"}
{"id": "train_070421", "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\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "989ee7acbba20193ed5089bbd58798e14f10f824"}
{"id": "train_070422", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "40905b1c088c3de9ee2ffd118ceabc9c768e016d"}
{"id": "train_070423", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "74bfb5c1b69ecac24cb5c916f03bdfca01dcfc4a"}
{"id": "train_070424", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "74b0d5e397d5464da1fb2244539aa0c9e3f3d6c5"}
{"id": "train_070425", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "ae240aab0c911304ce70da4e235d78c39cae02ee"}
{"id": "train_070426", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "94823a671a4affabea7f494e206061cd3e3f998a"}
{"id": "train_070427", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "65db80439e85fc7c113e93c8a1f3e02ea11b661b"}
{"id": "train_070428", "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\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "c81b815f6f769fdc70d3494975076d7dc231daf5"}
{"id": "train_070429", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["performance", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "79bd784ed9cb863781889d7b56263f7de1607fb1"}
{"id": "train_070430", "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\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "e1d247b3d602ce364d793850e29f73981bde58b6"}
{"id": "train_070431", "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\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "5c97609ab66e52f11303e4e269a3f4ed2b184ff3"}
{"id": "train_070432", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "410e5b62c71a500c301db04bd4dbb018a59efd93"}
{"id": "train_070433", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "d51a267f329a133c4e5d2c69045b6ddc4659c599"}
{"id": "train_070434", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "3343b62d848c0f9c3ba90789975e6cede84df78c"}
{"id": "train_070435", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "6622d0cda615719625bc98e33aa34fe794cf264a"}
{"id": "train_070436", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "812b4ffeedad3d3207a75c2ab59fd8e683f9c576"}
{"id": "train_070437", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "50db4a4080827dbbaa4ee38b8432d5ce3141f6e7"}
{"id": "train_070438", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "8f233417dd449cae54aa135b9e10d38ed9ead24e"}
{"id": "train_070439", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "7a8b5beeb1063f6f2e6e06a9a178ef435676d10d"}
{"id": "train_070440", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "5a5ac1489e4f475ac79c145ab35506e94982254d"}
{"id": "train_070441", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "6f2cf590cfcd313552a7612fae4c67a359ad2ce7"}
{"id": "train_070442", "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\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "c85f6326d644e1cd631a5c1a6e180ad160a24729"}
{"id": "train_070443", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a09f5498f5ecdbe10d64ba7400eeb3ef2ca3bbd4"}
{"id": "train_070444", "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\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c9b8325254fd79e3b0f08a0557792dc093c43cc2"}
{"id": "train_070445", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "779506a44b6ea1a5a857b901365dd715f994d4e3"}
{"id": "train_070446", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["performance", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "26c9d6f9b65665e50b9b89854473fd70491f959e"}
{"id": "train_070447", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["evaluation", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "aefd24804d67054802e3742a19f4796ae70f22e8"}
{"id": "train_070448", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "69c624b1e45658cb04eb8801d613386d72be7683"}
{"id": "train_070449", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "1a5d786a4aea51f905714e2101746c8549cf6f35"}
{"id": "train_070450", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "fea93faa173cf95a2713fcd3281d4a5fbd1c68db"}
{"id": "train_070451", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "d0fac72958b9437ad3c87f8cf75a4e66044c4eea"}
{"id": "train_070452", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "dbd96fc5177502ff8a466f50b1c613b065b23b35"}
{"id": "train_070453", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "c26fab5a069aaecda2fbf5ac5e97e164fa54dc36"}
{"id": "train_070454", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "325e0dc85a45f818918d4830147ae8da94badabb"}
{"id": "train_070455", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "da043f4b1cd88f8220c41204355ec5123164b739"}
{"id": "train_070456", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "d1f29069b3fba5e06fd9627e97da3681492f460b"}
{"id": "train_070457", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "7b409f42db3a02c66cb76ef1aa558eceb07cef2b"}
{"id": "train_070458", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "4fa501f6cddcf421831b4c0a6e07c64fb2d315fc"}
{"id": "train_070459", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "21dcbed354e352d0e5d96ab4e8039c3c866f8c64"}
{"id": "train_070460", "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\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "c42e3a0348f684fdc2504fac8a0f199d766a2377"}
{"id": "train_070461", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "eaa067419ce1a14d2cd19a27c3e4c73004cb104c"}
{"id": "train_070462", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "1d3ebfbe34e6340dd4b2799aa9c69b6ea4b9384f"}
{"id": "train_070463", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "ae32b9bae03142547db515d17f6d68aa63cffbeb"}
{"id": "train_070464", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "5f6c38dfc79485877a4e7c00af4a4c6ae5b46352"}
{"id": "train_070465", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "6c50a9daf8e3f849362d13450a5b87be6cbb59b1"}
{"id": "train_070466", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "dd129c8a47c9cbe7c4b669be48ce6e6bc5248a19"}
{"id": "train_070467", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "72b192b937b2e14cc36b84e8a2863ddeb0a4c5d5"}
{"id": "train_070468", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f00920b7aae7180356f3e671fbc2ca4e1f5c3bcc"}
{"id": "train_070469", "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\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "25e17c2d65f9697bc2f135265df9c19614fdcab5"}
{"id": "train_070470", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "301ac523373da6c37e620328fb57003463f40cd4"}
{"id": "train_070471", "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\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "401a4da398f05a96904b11551d147f19e8f852eb"}
{"id": "train_070472", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "abb6dbbd61e4846ab5bc05b4a5391f0f865c3871"}
{"id": "train_070473", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "cb3eb2679e6b9565e2924dbfe1fd844c435f0b90"}
{"id": "train_070474", "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\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "479c403360873701c0f7e4c3a65b08eb96f5665e"}
{"id": "train_070475", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "b75d3acd5a8fe167006098655e7fb38e29e38681"}
{"id": "train_070476", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "3d6ed9289b17a7b76b0c049327043070eb031881"}
{"id": "train_070477", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "f9f95a6633fa291cd0c4e873fdbfb8450eb479a7"}
{"id": "train_070478", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "b57d825150f3a930b5400f044d8d42d94a8ef0a7"}
{"id": "train_070479", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "c3a8caaa92c1d06561782ae9e33dd7b164cb2641"}
{"id": "train_070480", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "a3f7950a8dc33b5103121d1fc5f8cdcd2a555c68"}
{"id": "train_070481", "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\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "5207be653eea1d024d896207c4dc729d7c065dfe"}
{"id": "train_070482", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "fd5a9adef1acf599d4ab9f58a85250b2d3716132"}
{"id": "train_070483", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "26853e1d77ac754f38df9de3bf9ad2f3f5813830"}
{"id": "train_070484", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "348fe4e0e893334d269e3e4c9adf647b24814766"}
{"id": "train_070485", "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\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "beff1e9a4476f81eca8c75e8bd2bf3f07de4a192"}
{"id": "train_070486", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "b18c9c6a96c5b5444ab8099edb8a4c6d7425c9a7"}
{"id": "train_070487", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "d49213776f545b45bb8e67aceaaf0d88a11fcd70"}
{"id": "train_070488", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "63e8c951ca7f512054d911c3e02f66c2d3049a54"}
{"id": "train_070489", "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\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "a04afa1ecfabf146ff5d4dba70aba4dbf65d5802"}
{"id": "train_070490", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "12a4465bd90f2e2cbcddd9e397e612194985893e"}
{"id": "train_070491", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "ee933905944756175711744a5552640621c7d002"}
{"id": "train_070492", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "ae75c0dd65fa08695e54ed88f609e0fa3e4b60e1"}
{"id": "train_070493", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "7bc78deedde9ee1b92f547ede8c06ee5b11f6de2"}
{"id": "train_070494", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a084deca0468794fe11e04c0ec15788dcf99e40d"}
{"id": "train_070495", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "b18c9c6a96c5b5444ab8099edb8a4c6d7425c9a7"}
{"id": "train_070496", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "3928e43ed5968d2ee9b47bd55d972501da213549"}
{"id": "train_070497", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["agent", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "10f85a2aebc8480559d989334d4129339de2996a"}
{"id": "train_070498", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "7ab5d25ceacea14c2bdf4bda47f0a2bd1230d006"}
{"id": "train_070499", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "43adaffbc77603b53707c0f05a0f4a53da3b61c2"}
{"id": "train_070500", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "87e6eb4d532ea11afecda167594fcdb0bb22eec0"}
{"id": "train_070501", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "21abc306e9544bbea1fdcda2cb2c1060e3a04e0f"}
{"id": "train_070502", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "fcfb965520671cb1cd4ef622af6f6605f2573710"}
{"id": "train_070503", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "2c9e084d04ed7f9cabc27703dc43b4a6a1cb4fed"}
{"id": "train_070504", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "4c3d030ee7abd59074fbd3484a00d445efdf9756"}
{"id": "train_070505", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a246dfdf203eb52462f941376be434e3f40584f4"}
{"id": "train_070506", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b53eee8789f5ee2e335654258d955ec87c608a86"}
{"id": "train_070507", "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\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "c84aac87782ec8e8cfe2bcf4c8ff46b90e676178"}
{"id": "train_070508", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "670482ad4257ee4b3867e2900363359d08cdd025"}
{"id": "train_070509", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "5041307e52a23c842fb3d213b07e97e908e05ce1"}
{"id": "train_070510", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "160681b7fece773c1aac141625e90f78b035cf58"}
{"id": "train_070511", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "9ee1be780ae7ce43a7c19007fb57f969f105ede4"}
{"id": "train_070512", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "0a67f24bb224a5737b8cd9f0d428145617f6202a"}
{"id": "train_070513", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "474b249da981446a88e2a9f09e776553d4680557"}
{"id": "train_070514", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "c7455b8e188acb0cc050c446eb8429aeebcf39ac"}
{"id": "train_070515", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "15953359d4a5781ca3674d9bb7b592368ea22633"}
{"id": "train_070516", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "cee3b7f2e0bba2e54f65278fc16f3fd0d71ebae8"}
{"id": "train_070517", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "5a933a8831506f9f7e8aa756b369cde57e368915"}
{"id": "train_070518", "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\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "54f1ebd2b723ab225bf52fdc8114a06313a28cdf"}
{"id": "train_070519", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["performance", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "83b4d918c838f2ee8ee5cad82795f5613781a5b4"}
{"id": "train_070520", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "5705ef3eb4ad3a5941460535e813e238c10fa9a5"}
{"id": "train_070521", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "95be4481d88bcd0eaa80cf9b30b60eb2354be995"}
{"id": "train_070522", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "e1ca50fd7242047e5d8be1918b10219237c7d2dd"}
{"id": "train_070523", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "aa701258df97ec7ace0f28e8bc5e049df79a4369"}
{"id": "train_070524", "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\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "a4deda3c0e690f67d5ad837b13412fb48f438bcf"}
{"id": "train_070525", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "10277ebb3f4857d24cd9239d9ec0b0343a6b0377"}
{"id": "train_070526", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "e4c10beeffb6ac09a34873b45c221e897f96a46f"}
{"id": "train_070527", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "51353ca582ac261b78f3208a5161be1b70a3fd5b"}
{"id": "train_070528", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "f3245f4fd0088fab6fe14f27631978faf50591d9"}
{"id": "train_070529", "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\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "1b94a7850cce5419e1143edf8d609cdeff1a5d1d"}
{"id": "train_070530", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "8c3eb4e3c302fe58aae379ee9df4083616c04ab8"}
{"id": "train_070531", "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\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "ea5eb3584a03183b87c8711042666229948520b5"}
{"id": "train_070532", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "e933abd0e77f31c086e18011a9f04389bd41bff6"}
{"id": "train_070533", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "0023e937e42e8adbd0b27b79094e993986e673e5"}
{"id": "train_070534", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "2af146be4e68ef71b815eeb735f7bd8858a74511"}
{"id": "train_070535", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "0ba5c3024e6eb7f4bcac7c3c32480ccde5566a00"}
{"id": "train_070536", "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\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "de43703b0357be6f4f566ca3c9c4b9b1d37be96c"}
{"id": "train_070537", "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\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "d667fbd991d9dac5268b585a8e4fa0d9b9876550"}
{"id": "train_070538", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "7d8200883a0afff186b66a2fd56f56c81ae6057c"}
{"id": "train_070539", "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\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "8622291081792f8d3abf1bd8c779abc478c264f2"}
{"id": "train_070540", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "1974b0beb7d5c2b7c84887679406e85149fab38c"}
{"id": "train_070541", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "bf1a9fcfd0a736d84cc8574a1f43aa42e01ff225"}
{"id": "train_070542", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "fbef0d37adafa2dd4afafad11e305307612cf227"}
{"id": "train_070543", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "041604870159d6684ec69cb0bde6990e5bd49c16"}
{"id": "train_070544", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "577997bd35d08c4d1c406987e60f773b36551de3"}
{"id": "train_070545", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "89ccffa5c03a239e789e2bf9cc1335cafa4b1635"}
{"id": "train_070546", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "6ec0a6422410d5407d08f9b35580722cfd95379e"}
{"id": "train_070547", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "0ce50cddeab81461c385a67c944fca3b4321ed3a"}
{"id": "train_070548", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "bd831f3872a1e3bbbb4d768c780346840a99aecc"}
{"id": "train_070549", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "6db7bafe573591f667f8a18b024f682133529bdc"}
{"id": "train_070550", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "e21b7ce56aa6a51729a1c61ddaa8fc2e7cc51bb8"}
{"id": "train_070551", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "89c4966d30201abbead2807fe50695be427ba9be"}
{"id": "train_070552", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4112a0fbbd574836c8b549ceef67af59da9bb79d"}
{"id": "train_070553", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "087c974e6886d719175b77052f3486495b3acc15"}
{"id": "train_070554", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "aaf2bb1bd6678773333f842ebfd34d851a3ad47e"}
{"id": "train_070555", "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\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "2744499457e21f5d3666b4569333ce9226dd2b25"}
{"id": "train_070556", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "3026cc70c96576d4ae82493203b078012ba6c147"}
{"id": "train_070557", "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\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "6e3d896febab80cd934c838dfbebe420c63b94bf"}
{"id": "train_070558", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["data", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f6cb0a30188a07967894a0b6422d55895c5e4c6b"}
{"id": "train_070559", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "66da5d0ac2cab83810389406c590520ea8a38783"}
{"id": "train_070560", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "7f791ac4684c1092d0fc6365c79d0f4d97ec6de8"}
{"id": "train_070561", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "0fb9db30556fa12f004d8ffc7e3596f87aef3fdc"}
{"id": "train_070562", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "965b084038793ba67f14ebdef004a099ad589e0a"}
{"id": "train_070563", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["coding", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "a633f7a7ba9a079b8dc39e812b9764133dae6e9d"}
{"id": "train_070564", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "839a1c68f3962f767956c35bd9998391fffff1af"}
{"id": "train_070565", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "2f99146a6baf28ab3d60d35e89a550d56fb966ea"}
{"id": "train_070566", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "de61764b03b02bb557e59aedd63c09d3c9c674bb"}
{"id": "train_070567", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "095e142f3e3fbc06b2eedab7bfdcafbb0b3d1068"}
{"id": "train_070568", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "cdca2f1b9c93b8c81022fdcc4dca4fac44f09388"}
{"id": "train_070569", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "8548f043d3ce93d96d58633c08cbc3860ab72d11"}
{"id": "train_070570", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "a57e7c2a6c62a60a20bb604d6ef6f1270c2e2316"}
{"id": "train_070571", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "8403da2bc84ca312a5160d411a1ce4326c6ac147"}
{"id": "train_070572", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "ffc98aee4c97804b014101e24bc3d90805d89e0d"}
{"id": "train_070573", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "01ab69ba6b084f4a5165c969192a6f8087abbf36"}
{"id": "train_070574", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "512ee7e81e724af2a5d5560cef51c78960035cf5"}
{"id": "train_070575", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "4825df6d9557c2f7fdc73a5f4335144f7248d451"}
{"id": "train_070576", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "cd1db49fa51086f52e368e71278d7717d9eff71e"}
{"id": "train_070577", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "0690e629dd3fe259862a63dbb3cb156f6d63072e"}
{"id": "train_070578", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "bed1ffa2ad04866c8e1780982f2875d01e7fedb0"}
{"id": "train_070579", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "cc85ff1fbe43f00aa85d1fee4268e95c1552a5c1"}
{"id": "train_070580", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "794f72b6d022e8658a7b99b638a99108eba91948"}
{"id": "train_070581", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "002b78a88ec9bf73a89375f4eed771849a2acf42"}
{"id": "train_070582", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "0687a4a7d795585dcad6e9469abd2453ae18e679"}
{"id": "train_070583", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "3f9a1e209a34b3d273c5397481b9694e6006483a"}
{"id": "train_070584", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "c64b864a88e58291d7d60da2a13a8630f3de2c88"}
{"id": "train_070585", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["security", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "c596d29d903966f48bf418180b967e9a070798d6"}
{"id": "train_070586", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "f8622622280a5c62bf9a0cb63df1a0e0b4ebcae3"}
{"id": "train_070587", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "33e8e29742b8c3e387e422c291ace1c765fb13cd"}
{"id": "train_070588", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "b0fa9aa3e5685b7d23cf08ee5c054ced47581b02"}
{"id": "train_070589", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "d8cbd5a35bbf0f831b928ba3459654772561608d"}
{"id": "train_070590", "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\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "93019e1f95751743ee659aede41d825e237befb2"}
{"id": "train_070591", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "6be5a5b0d8d1ed4c9d556902c87cdedc6065350f"}
{"id": "train_070592", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e03e16aa7fc0752e9dd0e9329f3de91d4b0bc7db"}
{"id": "train_070593", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "66fae0a587584e14e6dad59a2a0abcbe4096c2cc"}
{"id": "train_070594", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "0a9dcd82adf6f9897f5357e6abd70a18a9aae2f0"}
{"id": "train_070595", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "3d42819932f7cf9723c2ebfe9f6c7b0d4453a209"}
{"id": "train_070596", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "114e2828f742477bce74698408d532a482dcac9b"}
{"id": "train_070597", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "cbc72d363eb41b38e595e01e0348adbe47f223fe"}
{"id": "train_070598", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["performance", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "6ff168e7ec6e1345e1af256d647fa3aea9be7c57"}
{"id": "train_070599", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "a37ad27321b6b1351200670ae663522bddb33609"}
{"id": "train_070600", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "877e8e3f36ad9738d9186b3051cfeaeebd55b275"}
{"id": "train_070601", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "2567d294ecb4ca483da707a11153dd23ed4f32c7"}
{"id": "train_070602", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "06538d880d55a63d451fbcb96fb16b755f9d7161"}
{"id": "train_070603", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "8e429f9584c33a8211a2d7a100dc352043fe12dd"}
{"id": "train_070604", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "a8deeff04311b85b9ba6469f967ca5cf4f9a59fc"}
{"id": "train_070605", "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\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "023a5e64585ea8a4a06d025a2be559850df1c570"}
{"id": "train_070606", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "0f3ea715a0118f4527a29661e7646c503e55ca3d"}
{"id": "train_070607", "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\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "21d61423ed92fdf147c0a98a1208c85a15d519d2"}
{"id": "train_070608", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "648a85ab7b8a6a189e3e684fe1eaf4957dfde675"}
{"id": "train_070609", "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\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "7b0f53cf307ce627c441ebe63668ec2b92185db7"}
{"id": "train_070610", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "c7c213df977d627fc1238781d6831664bf848e7d"}
{"id": "train_070611", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["performance", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "05b6893a0f9db588468ad3424981294393967f87"}
{"id": "train_070612", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "f7dee21b40709594319ed28ca0f3006690452963"}
{"id": "train_070613", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "44500a241d01ed4b386f2ff8948d7f670f36cf66"}
{"id": "train_070614", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "5b9f48bba08fee478fa5266a201b12d86e329f5b"}
{"id": "train_070615", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "65b4ad841c594f6adb032d8ea5bcf079ec325c6c"}
{"id": "train_070616", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "cd6be4723b4d9926eb5db3ee1bb0a75a799cb66e"}
{"id": "train_070617", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "baf86579c811b369754183c5c795d3669fc5d89f"}
{"id": "train_070618", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e5695c329fddcd4395c223d87ffb3b445f0b20f2"}
{"id": "train_070619", "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\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "9dc8a7702e949bcab039f18323ab5bf69b3f9c0b"}
{"id": "train_070620", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "21f57f818b82078b7b9066b97207d213d98edb27"}
{"id": "train_070621", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "42dd6a8d5d9c2c50380e1cf555c9d2e2c72139c5"}
{"id": "train_070622", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "d31c8265d72550686fa655e561e3864ad2481ad4"}
{"id": "train_070623", "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\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "bc88129c6909af849770a79492889507c3bde707"}
{"id": "train_070624", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "50e632ccb113e8b114bee92c64987f9420dc58b0"}
{"id": "train_070625", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["agent", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "558cf3b3ce3ee9224a122157aca7f6927ffa496d"}
{"id": "train_070626", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f269ad566856cf4447a2ffadfdc258677d70170d"}
{"id": "train_070627", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "c4be7395442f5bf168048ec30bcfafb2efc24d3a"}
{"id": "train_070628", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "4cadbab0570099395b823ddda2b2f7f67dfc5e71"}
{"id": "train_070629", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "16e5d790654ffde3832aa95738f605281b8e3ee4"}
{"id": "train_070630", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "31268fd362466d60ab889cdca46b8411373e20d2"}
{"id": "train_070631", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "debd69453fbbdf52670df3a228aec55c07ea7753"}
{"id": "train_070632", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "5a960a4b58e1ea0513c560cea0f653c929f26c84"}
{"id": "train_070633", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "062fd7e60e36ddce392659b4c9a8394768c37d0f"}
{"id": "train_070634", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "fd9e123267667bb25d0f48d5ff4152d100e1bb59"}
{"id": "train_070635", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "1811457b2ffa80955f88f837928dc4b01eeae3c2"}
{"id": "train_070636", "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\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "99ecaf8be1f905af9f95aaaee0bd970fae9ab02a"}
{"id": "train_070637", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "bd0717d676c03f6547c22071d24f334b64b9d163"}
{"id": "train_070638", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "002b413e98ef7589ee68f33e9fb5013fdd3f7568"}
{"id": "train_070639", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "23f92eae9ace48403d6c3d060138b92364074e71"}
{"id": "train_070640", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "12d72501c3b2948ac2d225a2f704fcd639186f7d"}
{"id": "train_070641", "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\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "b08b39c4d64d6b987472ad280f4bfd84e0b39caf"}
{"id": "train_070642", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "8991b31804113dc83ef669d1bf6495b88a75e20b"}
{"id": "train_070643", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "154ae6b5f256e0eab7abb81d8cb82f56392be224"}
{"id": "train_070644", "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\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "6a44646789349f01ebce65abc15eea70cdde27dd"}
{"id": "train_070645", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "ae259190a5cee6cb7c0be08fb8a5b6378268681f"}
{"id": "train_070646", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "d9e74981935432d5f0541cc4433242079b2b5130"}
{"id": "train_070647", "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\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "f84e00a4a7552e943c6cc94e1f5c3ab410709374"}
{"id": "train_070648", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "7ddd0c822d77b2f8e0bf179c3a735929b13d31b6"}
{"id": "train_070649", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "e828118d25a3a46b0f6008ed7524a737047df43e"}
{"id": "train_070650", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "f42f0c350677cbedba664d6d8cb31d65b43f4eb9"}
{"id": "train_070651", "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\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "2346b99f42bf2cd46400d4f273c7ae7bdce5d5b7"}
{"id": "train_070652", "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\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["governance", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "deaf0634c1b2e36ea78a260568027a6e4d8184d7"}
{"id": "train_070653", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "f62101a1032528617f0df1ca6acb89da57c48409"}
{"id": "train_070654", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "bc3b6e1dffc30f3d455cfd928a05769b61f27184"}
{"id": "train_070655", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "8ddf194f1d33ae7a3ecb339434db86056954e11f"}
{"id": "train_070656", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f4bddd687630474f9e5f1709e1630d3147aaf2ea"}
{"id": "train_070657", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["research", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "aca8c7c939727b7187241712b52853ca215b1268"}
{"id": "train_070658", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "775737078edbdab8936bb7f56f1a4caaaeb893b9"}
{"id": "train_070659", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "15f42d1872a2bce757774ff2b0b21688c006ce9f"}
{"id": "train_070660", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "92ef9db14135189dfde7d085cea273c712259339"}
{"id": "train_070661", "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\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "aa71b5ef3d35c6239e3414cf6a1044fb10011754"}
{"id": "train_070662", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "b3708c8064e13a002c35dd6ceaae29b7fb4854b4"}
{"id": "train_070663", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "0d39da68fa1650c5b422c165830deefd92c04bfd"}
{"id": "train_070664", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "bbf1e9d1623e274ad1405a22faa11678b515371f"}
{"id": "train_070665", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "0904d0dd93e05c6e9ea3b330c092892efff45c95"}
{"id": "train_070666", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "82632e59b5b5891f9fd38ab9f241b6d928fc8906"}
{"id": "train_070667", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "a4b7077f05f22d901a97f9bd9fcf4a583a991dfa"}
{"id": "train_070668", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "ac289dc75f73bfda237cd466378c808da4010546"}
{"id": "train_070669", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "8291e8167e2ee55ca4ff2c8e1634065467c852a4"}
{"id": "train_070670", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "0945cc6764131a92df86f654874d1b9829b76a1a"}
{"id": "train_070671", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "2a9d4835f9989093bc6ea0c82fb6c142003ff6d3"}
{"id": "train_070672", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "6ee7e3759d8211bac3ab79822cd0b81866d194b4"}
{"id": "train_070673", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "81061bc119596bb2a929c766e032069da8ec55dc"}
{"id": "train_070674", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "6c376405d100321d3904141d87d0b821e617ffad"}
{"id": "train_070675", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "9a5902f089ec1fe9723585faa0911fe39cf49ba1"}
{"id": "train_070676", "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\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["performance", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "feb889934b884a9df6c00d2aebafd283ac4f3b12"}
{"id": "train_070677", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "3249996b9df0448c36026ee17ee620e49b2e042e"}
{"id": "train_070678", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "90169fa78a4a64043829b2c533fc0f94aceaa354"}
{"id": "train_070679", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "843c29af8b6cbf8947ba905ac0a8181ba69c9d36"}
{"id": "train_070680", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "eaf7cc440439b9786cc081de38a3eb38e8904e84"}
{"id": "train_070681", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "1daed4e610284651e4b15de2231122e262bba60d"}
{"id": "train_070682", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "61d5e50e320a5131a20292355cecbf1d1524f3ed"}
{"id": "train_070683", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "1d166131c3394203deb6892df86d8f331f0b806b"}
{"id": "train_070684", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["security", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "686cc72761b0738ecfb569a80444adaf779cd506"}
{"id": "train_070685", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["security", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f76d8396975edfa5f862eaca700ff3499349ada2"}
{"id": "train_070686", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "6bc7bda13feba1e65833258bc9941917f0e497be"}
{"id": "train_070687", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "c154ebddbfa1718cf92abcbfe844d0c6984fd202"}
{"id": "train_070688", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "c2ff70bc1c141d48cd4ccbd5bf74a8028e2aa37c"}
{"id": "train_070689", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "5e0f4231d1d2232a6c1f8ac70017fe4ca40700f8"}
{"id": "train_070690", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "2865ea40b081e2cd1491e45ea076c8d3fff59594"}
{"id": "train_070691", "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\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "430116401fb346277a97f1daa4a45cf820a0466c"}
{"id": "train_070692", "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\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "ca8bc75aba2d5e72437db5e3c090111f57f9023b"}
{"id": "train_070693", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "9a11c8c62e74b0553e2bd2f2d06d1fa478730fbd"}
{"id": "train_070694", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "af2002ff5c86dec75f4a3abdf65a4d362d832842"}
{"id": "train_070695", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "d435f7183884d2c2b19a914fe372e1d6e1c042ca"}
{"id": "train_070696", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "8961118b61d6781ccde48162790ae04671879ee1"}
{"id": "train_070697", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "17f2daedfc6b831dc2042a5a94a3f4d2aef094fc"}
{"id": "train_070698", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "b03aaf80406afbd4f0c146f5378a3254091ef4a1"}
{"id": "train_070699", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "68e6866fffc444012eefa7487a86c3a3c282e3a9"}
{"id": "train_070700", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "31e054cb2661f654fd7edcae423967043968bc8f"}
{"id": "train_070701", "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\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "fe7d06e4fc885ba9c63a284f258f5ef24779c4e1"}
{"id": "train_070702", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "8ca780bd441c59244b5e19a875068946a8fe00dc"}
{"id": "train_070703", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "9097afc3633a1627747a90b2688c19c1acfa3722"}
{"id": "train_070704", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "af86376cc475b15935efebfd51f9e920f6434f56"}
{"id": "train_070705", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "039edb289f6fc3622228c6f084e0dd8c0db58296"}
{"id": "train_070706", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "b810319e8200a054e8a26329d859fadc2bf13d21"}
{"id": "train_070707", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "4e1f5821fc3d5dc19bb148e82b529072855ace33"}
{"id": "train_070708", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "45fad5d69c681df99342f89dd1cbf9c5208608cb"}
{"id": "train_070709", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "dfe1026f498e60f05f0b93c5984dbfbdff2c7e3e"}
{"id": "train_070710", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "020eb2282a54f79440757e44fc933556ebf9eca9"}
{"id": "train_070711", "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\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "b5de936e1783decf9280fea32355b8d933cb7d2c"}
{"id": "train_070712", "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\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["agent", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "f7c731e37880a7172ba486a969dccab1bd258967"}
{"id": "train_070713", "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\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "066236ad1eb2caf496321fc2dbbe1ac99216ff27"}
{"id": "train_070714", "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\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "27e32673cf36ccc5b627afd77d5de16939095ed6"}
{"id": "train_070715", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "213ee894f08e47a51a2f42d182e331d33ef835fc"}
{"id": "train_070716", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "f25ea302af469fbb15541f21f50ee461559fcde9"}
{"id": "train_070717", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "656ef0f0882babb4ed453edb65732b124013f17b"}
{"id": "train_070718", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["coding", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "f6f0e445d6c09b01c3837928161cea31f03f5830"}
{"id": "train_070719", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "90aeb892f9be8f0739478cff8ae790c163038c2b"}
{"id": "train_070720", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "966d03d39e8482a7e6932a476f022378878ea637"}
{"id": "train_070721", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "38f21e6228bbf8a38a68ce1b4b4c409dc4571fc4"}
{"id": "train_070722", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "1c38a7519a3adec37ee25656bae80df178679918"}
{"id": "train_070723", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "6f8df61cf2036018cd7775025d097543434b13a5"}
{"id": "train_070724", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "261c531ab1889169573af11d69f2cb95559ead2e"}
{"id": "train_070725", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "2605b84e2175dd2592aa2ed123176f1db4b0c8c8"}
{"id": "train_070726", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "11403b3a614ea3413d566708dc8a7e16e4579c95"}
{"id": "train_070727", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "4ee74b3f12a3d3536d0e2cd9943c9d85e57e4bed"}
{"id": "train_070728", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "c71a540bf385e4c3c56a5106bbc7ce9edce16694"}
{"id": "train_070729", "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\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "1b3ece31e5ce539d773f944501f16c7e587fe0bb"}
{"id": "train_070730", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "ee265008b4c8057ede4cc4f39777810105769603"}
{"id": "train_070731", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "cd1a082ba9209c3062c487b687653932a106577c"}
{"id": "train_070732", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "85c490cf7c3f0e5c326bec8f56b618fe526a6966"}
{"id": "train_070733", "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\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "27fac08ca6a87e2180147fcd3e7c318dd0dfa02c"}
{"id": "train_070734", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "9913852c96e08c292095181b215c0eac41d38a96"}
{"id": "train_070735", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "90e564ce9b818b2227e6f32515eeb54f9a0e782a"}
{"id": "train_070736", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "151cf90f9347ff99dc5e2fac7dcaf56ac62ace8d"}
{"id": "train_070737", "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\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "40ef3009b5770a6708123bc2382dd93a3d168553"}
{"id": "train_070738", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "89660ff9d53f74c8517f817241d4402449779594"}
{"id": "train_070739", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "1771a81b977ff611f12c004402e8640d8a87c34c"}
{"id": "train_070740", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "da6424ec33462e3e9d66c7d467bbae5bc36bfe82"}
{"id": "train_070741", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "70b0945c8a9af31f06326c6fa488cba399f9b51e"}
{"id": "train_070742", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "dffe0bbc840323f93990d3a33ceb53d8759e8a72"}
{"id": "train_070743", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "cc32cf05771b5de7a5296b3b5f2eff66f51d5e7e"}
{"id": "train_070744", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "ec86fbc9541c9d32fb80f8ca2ed28b54650c28bb"}
{"id": "train_070745", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "8318580888d3e8536b7a770ac70baf37139a8778"}
{"id": "train_070746", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "1d376aa991177beddf4b633745c8a029c64c2acd"}
{"id": "train_070747", "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\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "7b1a7bc70e925a2793c0059964dcc24705804cb5"}
{"id": "train_070748", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "84da7f8695b1f08444fd226fe76f4db13e3df34d"}
{"id": "train_070749", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "1a508431d185868361cc5c4d7571c73ccd03ee23"}
{"id": "train_070750", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "bb82482d5fed68a13df940e062045f72b245708d"}
{"id": "train_070751", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "72602667dd04f1b0c8ecd6e2fec4cd01907aad36"}
{"id": "train_070752", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "df5ec76d3f76ab131f738166d9b81d9efa7cc478"}
{"id": "train_070753", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "6a9bfa92458d830be232f40522c46cf2b0cf6698"}
{"id": "train_070754", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "0fd1c32b50bac40663e80394719c232fdf5327a5"}
{"id": "train_070755", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "9913852c96e08c292095181b215c0eac41d38a96"}
{"id": "train_070756", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "76f5afc613b0ce8a2daf1758330b16d023423c0d"}
{"id": "train_070757", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "934ceb158c877baf68d2be4e3e4c09c1ed8c4dfd"}
{"id": "train_070758", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "834ba8a8535cc230364f5f4678e38f1cb60cdce6"}
{"id": "train_070759", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "346f2887269d782207b15ebd39fd4f6e35cd35f2"}
{"id": "train_070760", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "7b0fc61f8ec5305f0e62dedff4fb7a170af3c2b7"}
{"id": "train_070761", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "cb6ff96f470954817af932b7b512c13e48740be0"}
{"id": "train_070762", "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\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "3d72623e91a7adc17cdb3786f46ee4dd165b09ad"}
{"id": "train_070763", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "fa46f1c2c84d9299bd1d38df8cfcc33e38d321fb"}
{"id": "train_070764", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "e7f38948b839bd7b2eb3d91ec0c9398510413f91"}
{"id": "train_070765", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "4c001b963af9ecde9d36065917db39e0ef7e5375"}
{"id": "train_070766", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "618eea9862917e056eb8bdbc902ed322a2454225"}
{"id": "train_070767", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "aba9bd0e72c15ca4ba9ea313a7a2d47f5798da61"}
{"id": "train_070768", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "b6a8e4b895985a527b92e323f2887359abfc43b0"}
{"id": "train_070769", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "a6d3c7ddebd6be396f9d4c8473f15e00b5b12d9e"}
{"id": "train_070770", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "4ee74b3f12a3d3536d0e2cd9943c9d85e57e4bed"}
{"id": "train_070771", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "99d13d10444739252a459b05fb4e5286a1530a51"}
{"id": "train_070772", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "ce607899df184d88e278bec5b77ed00a3456ac24"}
{"id": "train_070773", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "f54423c4c817f84b87c6f79e8df12f96b472513f"}
{"id": "train_070774", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "08d1ea526fbb5e6fdfbd1d02faf5ba0df45196c2"}
{"id": "train_070775", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "1b1dc26e6603f665e7b436a8bfdc11b2a00542d3"}
{"id": "train_070776", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "38225f08732a0d852d456643d533ff238207363b"}
{"id": "train_070777", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "0da8599a4a73437fabbfb88f661fad85b7f6402b"}
{"id": "train_070778", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "08f902af149364d9206cad1eb4ba10d1bfa27e08"}
{"id": "train_070779", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "d1d13a07ec3cb1dccc57be1026ffd44a98fff5f1"}
{"id": "train_070780", "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\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "fa187b39eb20deb39785fe15808cc50a4d38efec"}
{"id": "train_070781", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "9521e0fccebff97b661cd387954ff2fcad95375d"}
{"id": "train_070782", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "826f68390981e9644621dfa643e7d8f30d23af38"}
{"id": "train_070783", "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\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "41151031f3a7e6633d502a08214322bf64bda25c"}
{"id": "train_070784", "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\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "3feec814de6508fdb7fcbecf032a64e923cc96ea"}
{"id": "train_070785", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "094a8412e26f400f1fa42c2f4960319ed4abaedb"}
{"id": "train_070786", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "935224c1e0a97b0c353c325d2eee841d0cab864b"}
{"id": "train_070787", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "ee933905944756175711744a5552640621c7d002"}
{"id": "train_070788", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "d543348000df0380e247721b787b04ffbdaf42e5"}
{"id": "train_070789", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4987183b51be0baa8f785e27e00c668705acfc41"}
{"id": "train_070790", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "1acc1809e87908d4f960c95c173f18f3cbee3af6"}
{"id": "train_070791", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["agent", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "25e5b5d6ee75ca875bf93f3f609be976c4a4018f"}
{"id": "train_070792", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "7b3bd8309778659882b98de6860c9cae0e67c67e"}
{"id": "train_070793", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "21e48c885efbb2e0da6d9d3774f42b68a3265f5c"}
{"id": "train_070794", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "7186fcbf2568c7a8bfc500562af5cbeedc33e9ff"}
{"id": "train_070795", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "b598e1e6383594470af7410ee6b86524ec254f89"}
{"id": "train_070796", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "7324d52dde97ab46fa9f1465467f3c981a78789b"}
{"id": "train_070797", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "53804991b1642514958faed52d7d0429fab1cb4a"}
{"id": "train_070798", "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\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "8b9ed38c581eb16ab82dbf5199e665410167cd4a"}
{"id": "train_070799", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "2a0f09b4090a07a9f513a51d1230bb418b96a9e9"}
{"id": "train_070800", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "8706b13edfc8649878b6d321b812c6df0873753c"}
{"id": "train_070801", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "e2c97dec8c51913a09efe573e729a8ae4691ecd4"}
{"id": "train_070802", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "fa60158e5fd2dd17741b311d2f08db858bf43fed"}
{"id": "train_070803", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "79046f5e81eaca49b9b4ec0f837d891849bda8df"}
{"id": "train_070804", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "b996a8bb338ffd79f74a59a61ee20ea88a7b42ca"}
{"id": "train_070805", "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\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "8ad29fa8fd1417a4b1f703ff78a907b74aee3420"}
{"id": "train_070806", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e1ca50fd7242047e5d8be1918b10219237c7d2dd"}
{"id": "train_070807", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "13dec38d420f7ddb7759dd437aeefde604f765fc"}
{"id": "train_070808", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "90abe3c34e80ff67d2dd27be164e5b1fb29dc12b"}
{"id": "train_070809", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "cc5baa7c3fed2a159517cd5acaa8e76e9cb40a26"}
{"id": "train_070810", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "71894cd5a36997af87b8ea42bbc288814c2b68c2"}
{"id": "train_070811", "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\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "eca7f9f546402669909688bace8deb3eb599176c"}
{"id": "train_070812", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "bfe2bdb085d3dc680e1f207289c4ea174d7fd121"}
{"id": "train_070813", "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\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "eb303b3d1d4b42be2c3314000eddacacd0f84688"}
{"id": "train_070814", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "590c8694dcb16d5262e9009aadf280d891322209"}
{"id": "train_070815", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "185b987cc7723820a626db33c901833dd2cabb13"}
{"id": "train_070816", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "4013265145ba93a56ad6890db0c5016d13e4c3c3"}
{"id": "train_070817", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "1ed2a6f6aa44a9849cb8c1f007cbbaf137e3e722"}
{"id": "train_070818", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "8d137bd5f20cfc94286b806584d386e033b00d18"}
{"id": "train_070819", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "1c227a9c9e04733deeac3be6e6ce7f666ba1cf29"}
{"id": "train_070820", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f0e2aa109e315cd07223dfebe050cfe5ce944521"}
{"id": "train_070821", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "23f92eae9ace48403d6c3d060138b92364074e71"}
{"id": "train_070822", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "839b2b211adc2c1c9088a047882ac7b9739e13ab"}
{"id": "train_070823", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "e71d64c1e5e9c76ca85b26383d08d834a568d107"}
{"id": "train_070824", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "4bf4feed70744ca0670757a48629a4c56a7cf7c3"}
{"id": "train_070825", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "dd035b1316333fbbf9fa41aa9e96b5803eae4d6e"}
{"id": "train_070826", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "bf5d4d796973dc1cdc39373f1ec4a69d0fdc896e"}
{"id": "train_070827", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "b65b5a4720d9a3d47b3b5a009e426cc029ac641a"}
{"id": "train_070828", "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\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "8dc6f285c848af6b9a4a5643578a9101d6f2f68f"}
{"id": "train_070829", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "6dc8a7933fb793ecb6e8720a48780ac683bd4c5a"}
{"id": "train_070830", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "e38e5ca70db525888fdfd62a482783e22379a0bd"}
{"id": "train_070831", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "9c439c236900f2252b11c187e2b1863a2cf760cd"}
{"id": "train_070832", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "121ddfd9b0f3a81e432f834a8c852a2333a6cb93"}
{"id": "train_070833", "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\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "b9449ae0ddbc54981932010a5e13b1de95460359"}
{"id": "train_070834", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["research", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "5919ddfb6d50c00b003cfbf75c5dbe323e7be8d9"}
{"id": "train_070835", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "583816e20ba411dd7b31dc86a7dc8460e2bcf685"}
{"id": "train_070836", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "d8197adfefacf7d4a285655e47242aff4f3cc049"}
{"id": "train_070837", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "e7633eca5654c241d15c2618053a0ff20c882b9a"}
{"id": "train_070838", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "3ea23bd34645e39dbbc2fce46a03eb1c18b799cd"}
{"id": "train_070839", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "673b4854214eefa6df902dc11d07f30fb9c3d097"}
{"id": "train_070840", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "46c113fba90a511dc2df35e115470327426409e5"}
{"id": "train_070841", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "e8151462a12de65673dc5a15d132f4e7372f73e0"}
{"id": "train_070842", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "39e0aa0752d6eaa4b4aad49d13aba2c48874b91a"}
{"id": "train_070843", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "6a0aec13635bcd75069d97640abaf66d564fd226"}
{"id": "train_070844", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "8c5b1bc2fbcdfc51dcea89d0818423d4bd89775c"}
{"id": "train_070845", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "6a26409f8424407560bdbc3cb028048f3c810d7d"}
{"id": "train_070846", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "c07ca4f5c50797cf7d46b0abe389795fb7d79132"}
{"id": "train_070847", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "b03fe971b2c46a50b219d7ae447c3c6cfb42533e"}
{"id": "train_070848", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "9c12f9b0d8d00dd24913f09023e636a6e768d3e4"}
{"id": "train_070849", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "cc2acf8b8e7949dabaff5258fdd3394449834fb3"}
{"id": "train_070850", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["governance", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "b067cedcee70b4f80e8b1409bc6ab6f110f846fe"}
{"id": "train_070851", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "10cf5f8fa3e3645893f7bf88cbdce8944c909a9a"}
{"id": "train_070852", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "38dcf1bd1d286859a41f9455b4f729d532339560"}
{"id": "train_070853", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "e9b31765ed5cf0f3cf8954f38cd8879a904fa981"}
{"id": "train_070854", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f6e9a8063a1e62b879c40cdb0e2d6f5c3d06d92a"}
{"id": "train_070855", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "31c18fef10b18d3cb5838999ed873cd85ae5bd32"}
{"id": "train_070856", "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\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "8958ade8f909a2b30187e99dd6207a6059bc2d69"}
{"id": "train_070857", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "fcd2aa462d81e2ede7bb8d2d47159fa0c0b5758d"}
{"id": "train_070858", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "44c351aa806d259fab79cdb5bd2fd38722bb2806"}
{"id": "train_070859", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "e87c9aaba1e91ce3f639f1ee38f9dca1a285cb92"}
{"id": "train_070860", "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\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "d713b201c0dcf6a6585e02d47d72db22ed805b8d"}
{"id": "train_070861", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "030a82835f3535a12d1e444c369db84da74c52f1"}
{"id": "train_070862", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "eda5b1ea4085f304ff2ed1d278f171888fb319b1"}
{"id": "train_070863", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "4d630552eb4eb359de5dcc7f8ae1bd4bfa353a23"}
{"id": "train_070864", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "72b9a52abed45d8a5b099188a415cb42a4444c16"}
{"id": "train_070865", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "c31272b4c0920560c668b5564281c125bf826bfa"}
{"id": "train_070866", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "ae2f83e0a453f3ea93710a5cc384257cbc61cff6"}
{"id": "train_070867", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "66df9f989e1a5065439e599c7e060b000287c139"}
{"id": "train_070868", "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\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "79f1498c523e03656302ae86fcc6d5c5eee5fea8"}
{"id": "train_070869", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "716162ab83510e379a132b06f713e31b22b29778"}
{"id": "train_070870", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "19a3ea84b3a20e0053d09900eda5630968674e85"}
{"id": "train_070871", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "b022fe098ee897de2784d9f9ee90e52fa7d0c02e"}
{"id": "train_070872", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "097f1b175cc7aac816120becbde62f70f8909fa4"}
{"id": "train_070873", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "137cd941bb136326fd0ddca21436daf70ae95e01"}
{"id": "train_070874", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "77771b1609aaad4d4687bc9de23ef108ab469ad1"}
{"id": "train_070875", "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\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "cb073957286410ee6e934895c28f031cfafe2b83"}
{"id": "train_070876", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "b7052c3a690b9bdf618d8b56d1f3b42cda80b22c"}
{"id": "train_070877", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "926fbfe2af32eb82a6ecfa3e131074af63a52597"}
{"id": "train_070878", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "f087d8d109c7398c4cfea06a091a90362a7e67a1"}
{"id": "train_070879", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "6645062783ce12bfe045327c5d62f63a3352023f"}
{"id": "train_070880", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a81fdb7cbb37d53cf816597781881405118d5b33"}
{"id": "train_070881", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "b5294d4ff15dbeb88cf73d6dc43d04ccdef16389"}
{"id": "train_070882", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "e9dda9eb3be50be59547ec8918bc299a9f4c3481"}
{"id": "train_070883", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "2df37e12f2e57d0c980ce8ee65097def3b181ae3"}
{"id": "train_070884", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "ccc9a9f6f493323a9bfb7d3c54235258244ae0ee"}
{"id": "train_070885", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["data", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "50e867c9f2dc4d17a4ccac902b14858afbad149f"}
{"id": "train_070886", "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\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "302ef3bfe76efc1ad67a3886cab9caa5dd0c3dc6"}
{"id": "train_070887", "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\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "69611e544692ce9c8471e58540f240b7716d4298"}
{"id": "train_070888", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "023a8924ab224fd020fd893fb0ee7fa7e1ea2e91"}
{"id": "train_070889", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "ded28fa90fcff226196d987672461e272d439f08"}
{"id": "train_070890", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "eb9b64cc020624cd683bebcb47cbefab70bcae8d"}
{"id": "train_070891", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "03a418c3e6e4d04cdce139ca797cba3cee3b7693"}
{"id": "train_070892", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "a26d367c085e0c10a7cfe4dcf015282567b1ee21"}
{"id": "train_070893", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "d83e7f88b548409404a6d419cb14a1c411f82dee"}
{"id": "train_070894", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "cf40b19bf0caf7f8ae33d64bf2cc546ea4cc2e5e"}
{"id": "train_070895", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "91f59eb0252b388715e1247d23a9acab4fa3d89c"}
{"id": "train_070896", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "a17b4b64bdd8ddbe1408330b43a3ab59db74dbe0"}
{"id": "train_070897", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "70f3133ac490354341e2f5f029cc13eee20bfe6d"}
{"id": "train_070898", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["research", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "dbf61a5724746f0ccdcc375bb856c3212c5d6671"}
{"id": "train_070899", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "83ce991d5dd8baf597741a5d1ed22089433bb271"}
{"id": "train_070900", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a75f4854534bd9eb911335ca927fa1768c8fe22a"}
{"id": "train_070901", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "a7b6692bb60d639e4c501b95fd7dae36e8cc218b"}
{"id": "train_070902", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "d45db4bc6395e3668f4ec585636d829fc184112d"}
{"id": "train_070903", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "1804e522f1e7b425307b510013652647318433c1"}
{"id": "train_070904", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "8c4c8e7000d98e38ae32883ab5b24c4fbc89f689"}
{"id": "train_070905", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "c2d94d8fa4480945a54208f3d7db3761fd49436d"}
{"id": "train_070906", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "851833a4894c07e329fad289ed3b3568dc200e47"}
{"id": "train_070907", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "ce9f4ed6fb088b9757e5d9a6a5f3889bf7ae0ac9"}
{"id": "train_070908", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c05791bbc926e75c752f5436be861334f70ce93a"}
{"id": "train_070909", "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\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "3137be50a0ca59c148a21415872caef86a766e03"}
{"id": "train_070910", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "9dc75df125834872a15d79bed328f0f4bcd1e816"}
{"id": "train_070911", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "66f51f2946c149779630259b17e051877b7cd4e5"}
{"id": "train_070912", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "c9df823388d8bada9af6c5f2024cce6f48cc2798"}
{"id": "train_070913", "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\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "7efde6ccbd972356e4231ea118f5c5ea36d924ba"}
{"id": "train_070914", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "6f33df13250596c6507acd6778140e504ffaa569"}
{"id": "train_070915", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "411482e3e2d59e1dbcd9fc1c49149ce07440fe60"}
{"id": "train_070916", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "1942843aa4254146e6f66d0623030b8267efdcc4"}
{"id": "train_070917", "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\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "782e34eba2f30cc8096770f8ef32d4185df768c2"}
{"id": "train_070918", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "8ce48b8ec65ae1c25a2f7246564eef11be8a2872"}
{"id": "train_070919", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "54f03bcd2212ebf9cbbe291acd9af9d7cbcf817a"}
{"id": "train_070920", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "857ec5be3c888f4204ebc2bc50e1fec3e73fca8b"}
{"id": "train_070921", "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\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "28ad28fc46c70204e6463100ea47c0666236e4d5"}
{"id": "train_070922", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "84b9dd50382016dba128f526c743f39b669acb2a"}
{"id": "train_070923", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "e75b9c8957cf2a544567f6130f68ab3c1bf4bcbf"}
{"id": "train_070924", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "fa5b0f2b31dcdbaf8afe2a2b3931ebb09a4fcf4d"}
{"id": "train_070925", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "106be804f4f77cdd9a41bae83d8908183cfd35e3"}
{"id": "train_070926", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "7f7194a6f43fb36c7883cf133fac4168b8658fb9"}
{"id": "train_070927", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "52bfc84d92a70824450ccc3f6883be49225d2379"}
{"id": "train_070928", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "737b25f41e45e68d32a5659c05273a0facc5f42b"}
{"id": "train_070929", "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\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "6f5419bb6b4c7613d8130cdd52dcc08c122c27b6"}
{"id": "train_070930", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["agent", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "46b4b5bdc38bc0b575ed98a8a847e59b9c55be65"}
{"id": "train_070931", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "05c6ee49da93b2175ebe995fa135d43bb1a84fc8"}
{"id": "train_070932", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "17144cdbb9491e26e6d200a4c8ce428ce012edf7"}
{"id": "train_070933", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "fcfe53791b9be96720839478215a94cc3c5c3615"}
{"id": "train_070934", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["evaluation", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "6011bf2d8a331baf555d63b68b7ebfc38d8a7bd0"}
{"id": "train_070935", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "cd92fb80ee3f1abb1fb8fccf6f89f985fb3af129"}
{"id": "train_070936", "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\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "178548142c851fc6febb65768198d94b6d69fc87"}
{"id": "train_070937", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "4ab3de277673e6712a07c7e8eb396a5b8bd95972"}
{"id": "train_070938", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "3e5a534108e488bbb0264c63425ae802a39c1441"}
{"id": "train_070939", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "2bf5e05cef1aa1c5a3e27df4e9959deb4b301df2"}
{"id": "train_070940", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "714fcbb7f555f85e40bc701d5b73e1c73a4bc23a"}
{"id": "train_070941", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "cd9964289377d62866b020f4afabeaab72a667fa"}
{"id": "train_070942", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "3afb25b3747afc3c47ee7e333fbcef21bc8ce7ab"}
{"id": "train_070943", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "077da710bd4826078fe71ae6ad0d16a39ea95017"}
{"id": "train_070944", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "9d155b22dfdd2eec662fa600ffe7e6d485928799"}
{"id": "train_070945", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "5b633c1c9451d60faeb2541ffdbcd1dfc59d459d"}
{"id": "train_070946", "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\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "74d3a0caec59351db5d136f2763025f8e9a6d563"}
{"id": "train_070947", "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\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f93eaa471a7e4b0692b1830c206192b339ea08eb"}
{"id": "train_070948", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "4039bbf36da67fe7901eeb1095072bb39792ff83"}
{"id": "train_070949", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "a34087eac7d7c564c905babbd28b50ea37d20a25"}
{"id": "train_070950", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "105e7b7955a645c8aa57e3f367f879544ad24cb1"}
{"id": "train_070951", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "e3e62c6e31344b54cb0405a8798d4e9b6dc5dce1"}
{"id": "train_070952", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "7ae89a5ecfe9bf3c84cbe952bd5d8390358a0b51"}
{"id": "train_070953", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "db9e122e36f94ed412b8a8278b5c05419fbafb1e"}
{"id": "train_070954", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "001b9390b2c3d857be4fca63ef9e297d4298e32b"}
{"id": "train_070955", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a719a962417c8492db60fe908c507396858b27ee"}
{"id": "train_070956", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "7a8b5beeb1063f6f2e6e06a9a178ef435676d10d"}
{"id": "train_070957", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["security", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "0ad540b12966c20248420eb9e63082f25ff3b843"}
{"id": "train_070958", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "7e6eca6d992c772471376bbfc1c82ed9ff8c38bb"}
{"id": "train_070959", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "2e0661cebe25580b2c0a39aac4db07185b1d960f"}
{"id": "train_070960", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e1814be69088ea1fed092e28b16c8b8a7b523924"}
{"id": "train_070961", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "b6bad8065fc4319ca5024563a9f172643f4e0796"}
{"id": "train_070962", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "af59f67bfb38ec0cc4ce73a47d51b33b2e56b90d"}
{"id": "train_070963", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "3298760228bbe4f94eabb0a212b34992166bfafa"}
{"id": "train_070964", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "097f1b175cc7aac816120becbde62f70f8909fa4"}
{"id": "train_070965", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "816e3854ca77629aa9cc07cffa8138d4bb1525c1"}
{"id": "train_070966", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "a4b85ffd4a7437392f80d3dad234c28a046ec328"}
{"id": "train_070967", "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\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "a803b3cc280d03bbfbc7b0dc26756a52c577f40c"}
{"id": "train_070968", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["research", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "015748858be2e2328336c4504bd9588e43a9cdd9"}
{"id": "train_070969", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4e46300eac24ab803b0cc6a632d2af4fd08e0be5"}
{"id": "train_070970", "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\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "9c25bdbea83f1e693211d9649f81c3b5012bd7f9"}
{"id": "train_070971", "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\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "4ce6d79f1233c2e123adef635731e17ad28b8300"}
{"id": "train_070972", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "f2f0fc4497341f1d95826d47b33da2da6c0883d0"}
{"id": "train_070973", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["research", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a719f0380a9b3479aec45f97dc52f43d6f9e1307"}
{"id": "train_070974", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "3a27e9631cb7951760117339cd1ea3517d51a4b4"}
{"id": "train_070975", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "0893e2beeb08a9785ee78602535b6e018921bb0f"}
{"id": "train_070976", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "3847e97873e92635cca10ea57018ab258ee5ba3c"}
{"id": "train_070977", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "cb9ae3d94f1634f1dc33cb6ad0daa0d37fe36880"}
{"id": "train_070978", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "24d5d6ce03ed97bbb5def017c37c064cde21f074"}
{"id": "train_070979", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "cbd612e6c6caf25940fb3baa90b9f2af24955d35"}
{"id": "train_070980", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "47cbb81512dafa3baded5d011f045873740b984f"}
{"id": "train_070981", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "2ad4ec06895467646250ab1cc160c7996bb09ec6"}
{"id": "train_070982", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "49ac8d318fb4e5f672c5f198e43630da3c2ecf5b"}
{"id": "train_070983", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "af07e2e0f03b4c69c34dcb6e673e861e83a99db2"}
{"id": "train_070984", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "8c9331b91f76a9aeae741c819c8dbbdb92cab558"}
{"id": "train_070985", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "ed06ce600ab5bc678e3180ed50926bd1270b4bfe"}
{"id": "train_070986", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "b3bf61efcb7088b6d8a8f4ceb562a3120e525aca"}
{"id": "train_070987", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "9648e073b9d856563342dda305dea634476c8bef"}
{"id": "train_070988", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "505762040219f1659b725a342e8ce863879da074"}
{"id": "train_070989", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "1c0bb18311f6fd6e6399496b77b79e8f151077a8"}
{"id": "train_070990", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "5fb254dc96de9017db59092a2fbff43001baca95"}
{"id": "train_070991", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "f59700582e29128addc4d49b69549b1e90bdc95c"}
{"id": "train_070992", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "05676b04bd270892109662a4555b65448b444eb9"}
{"id": "train_070993", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "c8bab40ebeafb055003d45c217fdaf9886ca3fa7"}
{"id": "train_070994", "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\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "42fea324ccea39febeee5f4ed9c510f134694e3c"}
{"id": "train_070995", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "0ddc60fc61864d3962bcdec4e8c88e263fc8adc1"}
{"id": "train_070996", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["governance", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "7822553c4e576513d4f5940cb50500f42e84ad6d"}
{"id": "train_070997", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "8854f862f7115899a291e925fcbe0fc9118ea7f6"}
{"id": "train_070998", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "c5fa512ff6fb50f1cad46f9483c03c7963991ac8"}
{"id": "train_070999", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "c343390438a6beab0890526f1602f35e96ca46b6"}
{"id": "train_071000", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "26843346afc489848aebbebece5faf6e0a3b8c79"}
{"id": "train_071001", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "26e8c01c43f4895dbef441014ab31d01bc8f8e58"}
{"id": "train_071002", "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\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "dedb70fe20d68327caf36d505145a4b00966f81a"}
{"id": "train_071003", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "825a2aecb2ece9de399d80ed8c7f8a3839bd66f7"}
{"id": "train_071004", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "b235915d2f861ce74cdabdfe0a0c0af98b73fbc8"}
{"id": "train_071005", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a2f2f5a0aeb3c3b9843c23a0ce1f2679f990068d"}
{"id": "train_071006", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "cba1f35d697855e3ea6a012a87ae4dcc36c77f1c"}
{"id": "train_071007", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "ae6319b29c3d21b0d9e25048fded357f17e24b48"}
{"id": "train_071008", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "f9c9002ea90aba6a480c4b6f9a27b9b4c35941a0"}
{"id": "train_071009", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "a93804107e851d56294f01f7f75cb5a90fd60c60"}
{"id": "train_071010", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "63b508a3725dc3c5219bebd23d5c593b2f4e826f"}
{"id": "train_071011", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "da9deb67d7c8176a6759b66a87a23a806d9f298b"}
{"id": "train_071012", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "1abe2b2c056152d8d4ba6a23c651e76b383593b9"}
{"id": "train_071013", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "4f030bc84dcb0443bdf8be15647eb8171bd32562"}
{"id": "train_071014", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["performance", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "fd5ffdd9bddb0d6253803cf5a170059b96a75cab"}
{"id": "train_071015", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "cc4344cbe754089b59df478339a0360f23533668"}
{"id": "train_071016", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "41aed948f079bac2f44fedec68007dfcec63ffae"}
{"id": "train_071017", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "8278639263a3a3f9975c295584315cf4a225c6f6"}
{"id": "train_071018", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "0413991dc831166300c2616650ae0081f5230756"}
{"id": "train_071019", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "7ea2d7d0422ca00491a875f0be9b956b5d3504a3"}
{"id": "train_071020", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a5e058767b9a5f2ab4aeec746bdc1a3a49f7183b"}
{"id": "train_071021", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "0f887a4eaf20ee703ff88e6192bebc08644a94b1"}
{"id": "train_071022", "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\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "944b165a8ccbd9d2e75d47854c09613f10bd679c"}
{"id": "train_071023", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "58e68b4f04894799acb395f27c4455d11178b9ba"}
{"id": "train_071024", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "3bbc2c9ad1c93fb9a05901f34f540733611c8b3e"}
{"id": "train_071025", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "102315d009e93378d87a7cb5a4077f8432290f95"}
{"id": "train_071026", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "d571461bf0b5de7fa7d8207d0cc2a295d35dd8fe"}
{"id": "train_071027", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "9dfd330b5595c10aeb933e826a1a0c753a29f593"}
{"id": "train_071028", "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\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "751a179c9ded0f9d5b7b9b8c5ec96d4d0cbfc3eb"}
{"id": "train_071029", "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\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "582d36b8d363c144b7a6a5eb055e7982053c7a1a"}
{"id": "train_071030", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "ce75a1c46cae76df0549bd4e237bb6d199b8d7f5"}
{"id": "train_071031", "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\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "c96ba243a8852f67fef5d2d6d74d8d9b6d025f20"}
{"id": "train_071032", "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\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "27af865359880c5cca6022bdfd42c346f6e25f6e"}
{"id": "train_071033", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "dc257fdea2a2fb42d63a3e8b96db4339a473bfe2"}
{"id": "train_071034", "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\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "71b4d3273d604ce16d2b2c4a1b22e9bed86985e0"}
{"id": "train_071035", "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\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "b5ba44862c5a40ea16f0bd14f106ea3e07bee0ab"}
{"id": "train_071036", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "486c4853c8bb7ac48756df7375dc6a57a75b5f5b"}
{"id": "train_071037", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a962c80a73eed5aff7a0f235b1976a60814abd2d"}
{"id": "train_071038", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "d8ae3168be15c6d8be50c93c21710c8720bca542"}
{"id": "train_071039", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "c6682eb15e4cf9e794337a7f3471375973a26f92"}
{"id": "train_071040", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "dd176b5be9498fd1436d2d942a0c72c9b36b3f13"}
{"id": "train_071041", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c7f342b1bfee9c0f26c10d07e837b31f93033bd7"}
{"id": "train_071042", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "2f0425f1bd64ab2dbdb3d0364eafec8a59a1f223"}
{"id": "train_071043", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "2c87e357350b91fe09c0bb63379e0204f9712063"}
{"id": "train_071044", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "cd00b9a27e67455f45f5ba813678807cd1f7b214"}
{"id": "train_071045", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "0f51bfc252a6ea31ee9c279623473807d0d119d0"}
{"id": "train_071046", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "cb705017584ef1217fff07034b10efed623c3810"}
{"id": "train_071047", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "1104df2e56331cb7946622976d6bc11cbce89560"}
{"id": "train_071048", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "c21355a414834fa0d8f8eca7a4193a9d5ba20e6d"}
{"id": "train_071049", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "9e254152214d1b7f2bab0b82e9c1207d0d9f9a23"}
{"id": "train_071050", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "055f3c3f079724af7ff18b4914537f7ddb1d67bc"}
{"id": "train_071051", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "eb16fcba00b48bef219c287f8e9891983d219214"}
{"id": "train_071052", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "03a846664de842c11bc577767d061582b8682f93"}
{"id": "train_071053", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "63f21ccf3ed4ff215eee2eb2eed0243fc19562d3"}
{"id": "train_071054", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a179acb981927e01bbef94c0d0800b3d06f72780"}
{"id": "train_071055", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "dcd198fb4df67d401fa355f98262441c7b20d83c"}
{"id": "train_071056", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "4ed327dc06ae3b1ab6baab3d18ccb27a90bd7494"}
{"id": "train_071057", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "56375f0a38cb74434b6e21c23547ebb781f6e1de"}
{"id": "train_071058", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "9748eb6d6764ce39b8ee8aa70df104245a071b82"}
{"id": "train_071059", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "92638df62579a229e39eef6af8d07216d3260b45"}
{"id": "train_071060", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "f99b5aac1518c0b727c8e14bd198f7a874e9d4d2"}
{"id": "train_071061", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["security", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "ee82b41e5db34110f51f45198161098f76dc9c29"}
{"id": "train_071062", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "16e5d790654ffde3832aa95738f605281b8e3ee4"}
{"id": "train_071063", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "de8ea03978c26e95489e71e404dc743e6f7beb96"}
{"id": "train_071064", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "606733dbceb97a78ddfa399687e8f38e0f9ab2e2"}
{"id": "train_071065", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "dc34b2378f9b796e02cc1d6e1e780cb657b1224d"}
{"id": "train_071066", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f5575fc6818ac86d66168760f255454f5f163fcc"}
{"id": "train_071067", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "005cf6f07123af00845a56567dab8d5f996f1f3d"}
{"id": "train_071068", "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\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "20a9b0c5b4e517c484239537731732be0ad34672"}
{"id": "train_071069", "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\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "7459607ef0c53e3187e39d075b05249f920b161c"}
{"id": "train_071070", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "cc14326abeccc55ec91c499b7f3abe1bfed5bd2e"}
{"id": "train_071071", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "db998cbf63f233158924add6e9fc00aad233a801"}
{"id": "train_071072", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "1294e817e637a69e322f59e0bde6401404a357fe"}
{"id": "train_071073", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "b1d0ba0fe10fee4b9cea11f07c309a0cf2ba222a"}
{"id": "train_071074", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "445ba238370a97c70114a91bd8a518ba68a59668"}
{"id": "train_071075", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "b42e994c4995c69f772a063e63a4af94a9c7a3aa"}
{"id": "train_071076", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "5462fe50028b2608414dfce19c88038db93a847d"}
{"id": "train_071077", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "0eb64a143e854459490e1cdfb6c17f3c1744c2f9"}
{"id": "train_071078", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "ccd65bec1a55fe45a2e15a3e6cf2683bdffb5b0f"}
{"id": "train_071079", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "0dfafbac7fb3f52dcd57f77e65658b1ebca16d73"}
{"id": "train_071080", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "e6a4db9ca2ba432fb514627cefeb8b39a8cc68c1"}
{"id": "train_071081", "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\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "75b3bf5a397e3e570ffb80fdc0b6e14d4e554c1d"}
{"id": "train_071082", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "f319cc68dc3003ce57d6982ad6386f5abee876ef"}
{"id": "train_071083", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "95acdd83fd36b807e40464990e01c00bb191fa8f"}
{"id": "train_071084", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "b57b626e0ffa79d05143208acb65571dc6fafff8"}
{"id": "train_071085", "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\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "55be245127c25ae089d367edb076174319d7e76c"}
{"id": "train_071086", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "684c7f406d491ff79bd777d4d715882db2f66a48"}
{"id": "train_071087", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["security", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "c3299fbf8f9cc6eed3cfa210ba547fc142fe57d1"}
{"id": "train_071088", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "13c5d6524337a1b11d790cd01b43cc60cf5d5f77"}
{"id": "train_071089", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "d43387e6ef70a660f45a0cc449cff579c035f969"}
{"id": "train_071090", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "2ef06e1e90048d865e46381d692a6cb319980687"}
{"id": "train_071091", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "462184157ae2fefa9a2c74d6141e8496d629e4f2"}
{"id": "train_071092", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "52835234c5e6d800a442234bb3730360d97b02e8"}
{"id": "train_071093", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "f16e043d9e87e6693e054010ff1a341ddf51165e"}
{"id": "train_071094", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "2584524c73b69f93eff21e6f2a48ed29849a79de"}
{"id": "train_071095", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "b109f625d86d250cc020e3c88c2a6447a750ce79"}
{"id": "train_071096", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "ea615adfdc9c6ee3202725ba8f360f05fe23bdf2"}
{"id": "train_071097", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e2f97fe4338012755ef1d4ee0b52be89c49ef3a1"}
{"id": "train_071098", "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\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "725404e627728ace187e4e49295eb7f4073556a3"}
{"id": "train_071099", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "587d04af41e00de391dddb98b511f476d7a01ef5"}
{"id": "train_071100", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "9c11c9e5cb464f96222e0c583aa9fba9f85db801"}
{"id": "train_071101", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "9bd02ed242bdb4b71f47dd6d6eeef525489727a4"}
{"id": "train_071102", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["coding", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "2a9cd653846bf30d7124fde6fc8321d23cce63e0"}
{"id": "train_071103", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "2cb4d05a3dd791bc6be4fd22d4285fd0923be01d"}
{"id": "train_071104", "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\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "d73b93871d6cd222711a8833d1a8a4e82cea354e"}
{"id": "train_071105", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "c228268d55ae894870754688b3293a2ffb5f4b55"}
{"id": "train_071106", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "8942ec9e751e47aa43c303d64805d449ad458422"}
{"id": "train_071107", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "58274e7ceda56ec6f9c5c0b375baafaa374ed234"}
{"id": "train_071108", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "39ade47c9ab33873eb8b433c357df15e748a7e68"}
{"id": "train_071109", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "69ce7bda7e1f1afca9b8010192723205e0397cbc"}
{"id": "train_071110", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "fd12de49bd960f0cce55344e72bf94393f61c739"}
{"id": "train_071111", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "77c06836e6bed26bf6b3fc3e881b5f2cfccedb01"}
{"id": "train_071112", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "be0d58782dc9dcbacf19acc11479012781c47be2"}
{"id": "train_071113", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "6768839536164383bb6ad5213b96d85e56da7860"}
{"id": "train_071114", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "e9943bedea11c8d5922edff90d986ae3afede6eb"}
{"id": "train_071115", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "6011bf2d8a331baf555d63b68b7ebfc38d8a7bd0"}
{"id": "train_071116", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "a17b4b64bdd8ddbe1408330b43a3ab59db74dbe0"}
{"id": "train_071117", "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\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "3632d7b945e37744d2a42ddb24eb3820d08b55c8"}
{"id": "train_071118", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "a03e075ef356f884479755acea7b31217f2fc8e7"}
{"id": "train_071119", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e394d9de662e36203fa06d382d12bae0238fded5"}
{"id": "train_071120", "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\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "5d9a339e19659247916020cfa2675b87bde9ef97"}
{"id": "train_071121", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "ac2bbfdb44af3f03c1f7a6c46515f17249eb778e"}
{"id": "train_071122", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e1095cba53f9dca58b551d1380ef779a49cd9fe8"}
{"id": "train_071123", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "c8ba74f8d667208e1d61885e9e51e8092e5c3865"}
{"id": "train_071124", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["coding", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "e2b32dbc399152c7093cd34fbbe785ce215fe1cb"}
{"id": "train_071125", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "b862a66fd5f4719c1702f82b183afa221e7f60c8"}
{"id": "train_071126", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "9e66d32f015ad60500f5063f03536e006b513a30"}
{"id": "train_071127", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "7f29cac5703b8e51f56f140e458a7da405208472"}
{"id": "train_071128", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["agent", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "978431e2881c046f35653999bda7e943b7bb0360"}
{"id": "train_071129", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "63cd31b936571754b15af485b618c6a5c879654d"}
{"id": "train_071130", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "4666dae49c5f2868a0c872312b96419b673491fa"}
{"id": "train_071131", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "5eae9a2a5e984922a42b65aa92e46384f954053f"}
{"id": "train_071132", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "22e5322b20809d42fb090c25811b1416d0949d9e"}
{"id": "train_071133", "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\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "7aee87d911789afe1cdbcdc043539d165ef2260d"}
{"id": "train_071134", "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\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "c4457d9efe5e282ae3f5b23f454b69b72eb79543"}
{"id": "train_071135", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "16c4a34e95b33e21d87bb60cb5b8fa0adbccf7cc"}
{"id": "train_071136", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "eaf391488dc6e00b82ac540836a84162ddbd2ee7"}
{"id": "train_071137", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "1abf0c9657b4d32d82a9e1fb12b7eb46a8ea9b1f"}
{"id": "train_071138", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "5cac570a3fd2ee9d98bbf5f03d1e56560efa840b"}
{"id": "train_071139", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "d59cb81588525db967ef28ae084c770adf82f9b1"}
{"id": "train_071140", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "2226726b92c9e0677f91c6fd011802b762bf72df"}
{"id": "train_071141", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "fcae083f86a1479b42e8e18801b161a7eaf72745"}
{"id": "train_071142", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "8291e8167e2ee55ca4ff2c8e1634065467c852a4"}
{"id": "train_071143", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "62381e2beb8173b4c35166b8483bde1aea22ed1f"}
{"id": "train_071144", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "3f6da69d11c9d427ec30a77c14fb2d554ccca49b"}
{"id": "train_071145", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "5d2d3bd14e5e24f89e46a0763f302010ec96ace6"}
{"id": "train_071146", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "25d6353adc35801f4992e43bb3a35ff10fa51922"}
{"id": "train_071147", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "1e911462289e599ea446cfe91ecc8a6669c9f806"}
{"id": "train_071148", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "01a8f6a4f16a8ded6c921ba333458e1c0164d9df"}
{"id": "train_071149", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a8e968982663fcbf14699b954df55103d52c3f69"}
{"id": "train_071150", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "5ed76217140e65d3c494f24fc7f96dfa105220b6"}
{"id": "train_071151", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "a6abfb8f4225a3ac57f32bbda1a802280588ad90"}
{"id": "train_071152", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "28f51f91ba36985c2e833dc9b8e20bd0f6083f2b"}
{"id": "train_071153", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "455af4af9a9ba58a780df6796dc625c15688fd56"}
{"id": "train_071154", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "98fa5353bf2b2de9c3006907a0a1799e2a0b45c7"}
{"id": "train_071155", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "0bd479ec7248877fe5cfed84f2a4a43387b24d96"}
{"id": "train_071156", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "a7bd5d56ae09c512ce0cb36a3d32bec065798d81"}
{"id": "train_071157", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "aafffeaddf3fc200f0b98da47cfd83b27e0a35c6"}
{"id": "train_071158", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "11b2eaf000ae19dd385abd8b8d73134d485df5d0"}
{"id": "train_071159", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "27de4c8cebe74c351cdc7448f971c2ebe48260c2"}
{"id": "train_071160", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "830630f9764510ec95102f653be722fae77cce7d"}
{"id": "train_071161", "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\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "35af41e2429fa4c4e1e2e75b044a1c485d022fec"}
{"id": "train_071162", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a682b9967c8c6c232eb54056a60cb8fc34cdb26a"}
{"id": "train_071163", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "9094fa3e3ef2bedb873fb96702afc27a844dee5f"}
{"id": "train_071164", "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\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "2e288180c384d7e10e8d1882b4e4033430c06b96"}
{"id": "train_071165", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "86056c4caa7412abb7246293b985975235be18cf"}
{"id": "train_071166", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "7adf3c05e4f6fa1950e588955398c9f472fd0cc5"}
{"id": "train_071167", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "058312e60f300184acc79f768e2be5524670fa57"}
{"id": "train_071168", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "b55b3de30c5807cb4d3ea1c8d5844a4a8631d6a3"}
{"id": "train_071169", "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\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "875f5e5d42edb5ed9a8f01442f24836354989e66"}
{"id": "train_071170", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "add960a85c0b3b2de909503cf30ab05faf83f507"}
{"id": "train_071171", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "ec2418059dbc1dd82e5bda6e7c0b303844204b8d"}
{"id": "train_071172", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "6ded2e3290e9d072571d2455b4e2f9255a804558"}
{"id": "train_071173", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "cd5f845724fefa014580b3bb9956c3486c08fbe8"}
{"id": "train_071174", "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\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "cc6036d68f774d2b3fcbfc28bbe5c2821b9c90c4"}
{"id": "train_071175", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "cf91af3790dd2ae809e608ce6cf58d04531e4a86"}
{"id": "train_071176", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "8798f178322fc2f4184fe027ee8bffef5a61d170"}
{"id": "train_071177", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "d6d6c112f14e394f6e79a6d51c6a67b9b77624b7"}
{"id": "train_071178", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e297c3a7504e3e95b1256d5319bbb5ecfe80866c"}
{"id": "train_071179", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "51c38da93d44431ebd53cc4be20cffdae360b551"}
{"id": "train_071180", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "3298760228bbe4f94eabb0a212b34992166bfafa"}
{"id": "train_071181", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "13596348c92fc2ae976f83d0ca3f938e67c2f35a"}
{"id": "train_071182", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "8e0273ec161d7277b15db018b88aefa333297a20"}
{"id": "train_071183", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "1a3c6deef942d76df2cb22ce0c4d9afbe8c3dcec"}
{"id": "train_071184", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "fec61aebc519c1fb9b3706b880623cd077075dbd"}
{"id": "train_071185", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "3c2cb365191d799ee0ad7d325d8a9d7d46ff086f"}
{"id": "train_071186", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "89b869c1f5673fc5aaa205560de7bf06735afded"}
{"id": "train_071187", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["governance", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "5dfb467db2e487566689fcf97b57638146e5912e"}
{"id": "train_071188", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["evaluation", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "af4ea35e355aa9f030030bf170832e82c192c038"}
{"id": "train_071189", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "cefc8144dced5f7aad1fcd9a7189ddb856066266"}
{"id": "train_071190", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "c7a398fdf6f9aed23351d5ed57935fa01a2abb18"}
{"id": "train_071191", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "843dbed2f0274440416ef79452031bb94bcb0850"}
{"id": "train_071192", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a7e608f36ec16775b5d642df0b37eaaf9fcd1756"}
{"id": "train_071193", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "7f93e3617c9e556ec9dc72b4ad2d69c7a70485e3"}
{"id": "train_071194", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "a842f5d8b003f24b28203bdef5cbec7ffe0d0b12"}
{"id": "train_071195", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "559aae1a4015559ad7901d5da06c44c809930707"}
{"id": "train_071196", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "da0cd82555e9c25bb076b84a509a5adc83095151"}
{"id": "train_071197", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "ad60217dc60acfcdf6ea89366fbf93ec99509a65"}
{"id": "train_071198", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "661f74f3642eba5f273e079ec0bd467372adff76"}
{"id": "train_071199", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "e9b44e56cb8231b525fb2c1076799f3af73c2765"}
{"id": "train_071200", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "4abdbd5eaed653dde60ccd8a5de1bce2a188f706"}
{"id": "train_071201", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "92660c9a0f480b5515219e7fad5016e2693c4079"}
{"id": "train_071202", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "7adf3c05e4f6fa1950e588955398c9f472fd0cc5"}
{"id": "train_071203", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "1748cdc68f57df65008b8c89216704d9b0c59f70"}
{"id": "train_071204", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "fe8f8e804bb6df9b873ed0eb5e2ed2c74b908315"}
{"id": "train_071205", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "54d814c66330a115b26b6dd68faada3e2dee50aa"}
{"id": "train_071206", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "6b393496ca8fec8334ba2626e0b114c5f1741adf"}
{"id": "train_071207", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "3ea78ef44d7aeb27ccfd5f0e2c1a4ca6b5623c4c"}
{"id": "train_071208", "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\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c4025bc95de887432bd3cdab460a84c560e7c7ca"}
{"id": "train_071209", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "488d9bb80273342eab6ac9cf5999f25d4926c1a4"}
{"id": "train_071210", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "83a8b89c3f8b337fe1bedbb3faf273c336ee4015"}
{"id": "train_071211", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "c2f022fa82f18b728c1065f52c566adedd01b74d"}
{"id": "train_071212", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "2b99750dac57e46daf7d149c8c33ca443212130a"}
{"id": "train_071213", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "6e08e692bda81c92588e587e67e4b6c3e3045f3f"}
{"id": "train_071214", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "1ad37656e4e7280754148d8bf8a4ecd7a655cf5d"}
{"id": "train_071215", "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\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "b1d3404c5bc2a9c9780a4d413d7eb524eb8c4f38"}
{"id": "train_071216", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "5c8d884c5ee78c844a86f9a20d4ed72e9a91dbb6"}
{"id": "train_071217", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "3369cf6323387ab9135f703fd5585a5c464c91e1"}
{"id": "train_071218", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e8db5bbaaa8b1468d11e72fcd122614d8ff047b6"}
{"id": "train_071219", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "e9a5e2bc7029e9aadaeade7d45929bc4380c2100"}
{"id": "train_071220", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "5b042f5300b5a8cad9c832b8f54750e7b7dcc326"}
{"id": "train_071221", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c19189f159581c0901b85b72cf232ab480dd26de"}
{"id": "train_071222", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "8fc8d6616db2578eccb103b58a88af1c0c035e1b"}
{"id": "train_071223", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "99555d9eec57abb2f83daeae1a142dc931d6dbb4"}
{"id": "train_071224", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "1e668dfa99e36b274b5c1d839aa1a03ccf475ef4"}
{"id": "train_071225", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "003247c03aac727caa81619533398cc0ca24415c"}
{"id": "train_071226", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "aeb5b517c4d641a0f1f39344e5aedeb5b7b7639a"}
{"id": "train_071227", "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\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "b2cfadb1c247fefecf69d3c9ffdcbdbc259f4ebf"}
{"id": "train_071228", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "bb797de58e3029700686fe8ade4d3cddad2c6565"}
{"id": "train_071229", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "b32b4b87b3ea13db2e76c9dd13b517dccf6a3a98"}
{"id": "train_071230", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "ca19a86ceb074347ec497393f2cf5b058859abda"}
{"id": "train_071231", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "a15b5e3c3238b0f7839c8fbb0840b7daadd73fc6"}
{"id": "train_071232", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "e9371d78a44e8e671c34ba862483b26b3a6e7a1a"}
{"id": "train_071233", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "b697418c684f1817950d9b83150e35574a2ba690"}
{"id": "train_071234", "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\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "92d063abcb549bfc9e210b192ce7586d5fd03670"}
{"id": "train_071235", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "bd6ae94db13ec963814b2b58acc84ba35a3eb5ba"}
{"id": "train_071236", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "127dcd783cb41ae169cce9b6ae73b2ada4cfea7c"}
{"id": "train_071237", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "3d364663fa34b7603dee80e7b3c58c6ea21c1b57"}
{"id": "train_071238", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "45c8d503212f49b9bfcd75b0fa2a6defdaf6f1bc"}
{"id": "train_071239", "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\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "7363dce65639c3b3ecf8353c2c7aaeb5a4b8e0fb"}
{"id": "train_071240", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "819664cc942c63282bbafc54f8b7a1eac92172d2"}
{"id": "train_071241", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "175a9e4f319e89d6c8e3643d89d52692949bf0c1"}
{"id": "train_071242", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "cf9d8a803b5055967a1c16ac4ffd69367a3cd6b7"}
{"id": "train_071243", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "fe29695897cfc9a19a81cc1794ce1ab7b8d3026d"}
{"id": "train_071244", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "ef7a008f6191a9dd7f951bdf38b6c32bf9461620"}
{"id": "train_071245", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "9c3abbde1c5a68ed17d483089021521fff4ddcc7"}
{"id": "train_071246", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "ef4ee3c849430de8eefbc430653ffdc3d0cd200d"}
{"id": "train_071247", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "639823bdfc4ac8d8760bbc62d00df8da52d12d9c"}
{"id": "train_071248", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "b8a9fda8fa1e490435bc8675be75d53da8ac7e7c"}
{"id": "train_071249", "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\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "2db594e7c9eade79fd645afe068b48ac050493d8"}
{"id": "train_071250", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "0d54c1976e15efa9750d83cea0bdaeb96afe01ce"}
{"id": "train_071251", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "a979930e9dcc827e30f52503ab51432c63f20a97"}
{"id": "train_071252", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "82787b20d47a874d7f27caab355785e7c87b0d65"}
{"id": "train_071253", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "3a37a172a720639ba0536dc133585e7258fa4734"}
{"id": "train_071254", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "8dbd9a251137edfc333b6b0303f1fc884ff485f5"}
{"id": "train_071255", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "f77112c603e32c944183cdd73e0f462bd7a57b9d"}
{"id": "train_071256", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "ea79651e8dcf292f82b80b7a0dfa6295859a8de7"}
{"id": "train_071257", "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\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f3fe5e77e2ae2c961abcbe1b119c88a8686685a8"}
{"id": "train_071258", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "d7e02ccda21533a39e6b05d932de08220a96c123"}
{"id": "train_071259", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "7b1e9b3775af3cc7b96dd9cb5a15ffba35ae824f"}
{"id": "train_071260", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "00b5da9b17a91c4d170e3dffa9104e9628b662d5"}
{"id": "train_071261", "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\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "2ed845100c01ddc480b90dd93644c999e565af58"}
{"id": "train_071262", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "49fd27fd6a9669a94da1e83dbac1fae58d0530d2"}
{"id": "train_071263", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "9e110520e86fe6189f3d86c0c397c85bc4b9802b"}
{"id": "train_071264", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "7b5bda431c60556cacd7a5464fd3c3ef9223709e"}
{"id": "train_071265", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "90cbc8fc120db272ae5eca4838c2586affe0f91a"}
{"id": "train_071266", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "c74564c3a7dfa9ac660b5d16f3f58402cdcdf876"}
{"id": "train_071267", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "c32a9f1e74061a065c9434041589c0e9f869b578"}
{"id": "train_071268", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "d92cb5963fba48e8f5c5788b4504c5799868293d"}
{"id": "train_071269", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "a3ef192a82f1f87fc4d113d8350623eaf4e8490d"}
{"id": "train_071270", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "b89abe01ab60e71913d270cd268fea33f450e75b"}
{"id": "train_071271", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "b0dbd6a60eaa03d9a9b4e2e1c5a27aaac0c1fb53"}
{"id": "train_071272", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "884e1cf1cefc2187e4de73b208e89f48809250af"}
{"id": "train_071273", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "348d26f2687e34fef99ca2308057214dc194e560"}
{"id": "train_071274", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "8e8953b3b64855575688024ba99cd79f37815912"}
{"id": "train_071275", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "e08cdb80245537fb6998274590c5fc9bdf94308a"}
{"id": "train_071276", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "0ec3b46a455cab005edf3f0604fda09d2085d432"}
{"id": "train_071277", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "d76e7302e1a0733061149c9eae4b5732dc226ba6"}
{"id": "train_071278", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["security", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "ebcb3cd8f7f994fc60f0251892abda2ab3049400"}
{"id": "train_071279", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "e314cb0002660b0c79891ed7f5e2e93127ed8e07"}
{"id": "train_071280", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "b198047b54f0cd663c82625662119091be871a2e"}
{"id": "train_071281", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "f0115bb715da4b002da0683122ee9f8d0560ca41"}
{"id": "train_071282", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "c74564c3a7dfa9ac660b5d16f3f58402cdcdf876"}
{"id": "train_071283", "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\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "cc6036d68f774d2b3fcbfc28bbe5c2821b9c90c4"}
{"id": "train_071284", "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\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "ecb5e2cb57be29dd5b54bde3077b1e22cb04a178"}
{"id": "train_071285", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "887fcde6e63be1e969c9e0467fe29436b1f685a1"}
{"id": "train_071286", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a7e40c678eb26e0027006b1aca7b761c0eb5d073"}
{"id": "train_071287", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "3042a6e82ef2325163d55f2bba475f3f11d8a3c3"}
{"id": "train_071288", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "02927fd0819147b80470640a7d6ad7ffc406dcac"}
{"id": "train_071289", "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\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "66322ad4f646a58fde2a86bceb04bedaaca2e223"}
{"id": "train_071290", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "c23d55930d6963f172752017bd52edca7c7c0bad"}
{"id": "train_071291", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "435bfe5ba13aa9877a01dc589ab46ace720b62ab"}
{"id": "train_071292", "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\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "24c2ccd5cfe97460623e91e63339f34e1575ecfd"}
{"id": "train_071293", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "804a1419c8f91ed10d2209da795a7c17ac41b447"}
{"id": "train_071294", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["coding", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e0064d2eb8301fbe85cb858974990918c329450c"}
{"id": "train_071295", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "6dfbc89ce1acd687f4b4a4bb43faea8f4cc15cc4"}
{"id": "train_071296", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "912d919eb9a04d230530c8fd1d1e068c3e2f9754"}
{"id": "train_071297", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "6c804547d12284c0df859cdc3508fa022a154de8"}
{"id": "train_071298", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["agent", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "eaee2d0fd83cbdde646879e0ee1b2961d2ad2f67"}
{"id": "train_071299", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c1969ba3826379e5a9465077d854fd490e3a68dd"}
{"id": "train_071300", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "721d9a8481de72c44e9ed3affc65e69622deee06"}
{"id": "train_071301", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "e97c62f45f7ea94b46507d3e57b5764d6ca1784f"}
{"id": "train_071302", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "857ec5be3c888f4204ebc2bc50e1fec3e73fca8b"}
{"id": "train_071303", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "0340317aae39299daa5be761fc0ab926fbd1dcfb"}
{"id": "train_071304", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "9fcde85cf9d38c447328b68e5987edfccb36437d"}
{"id": "train_071305", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "fa996613ca411e20eaeb2ddd4a8d141791eafe80"}
{"id": "train_071306", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "173e439712a4eaafe7fc553c83fde12224219a39"}
{"id": "train_071307", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "bda23d4796af34313c42663104c18de7dec5127b"}
{"id": "train_071308", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "4ae32b39e316f24b98bd27f47d406b348ed9ff4a"}
{"id": "train_071309", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "7b25e8f5a65c7fa5d457e8a6809dddd9d96a6355"}
{"id": "train_071310", "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\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "41151031f3a7e6633d502a08214322bf64bda25c"}
{"id": "train_071311", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "fcc99ce06fb8dbed81021de35c52cb641f1a600d"}
{"id": "train_071312", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "14d06f97f59eeac520d3052fc13f69e9cd1e1305"}
{"id": "train_071313", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "d6ffcfaf1218a7ee7eb468311f8e54e9f99eaf61"}
{"id": "train_071314", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "6c4e353e6f8e469c552dd2190d9d710b44a3ab9c"}
{"id": "train_071315", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "144a7b9be20bdad3301344db78afd82b5a028234"}
{"id": "train_071316", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "d8212cb8a16afddf9d53c09d5928679ecdaf2058"}
{"id": "train_071317", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "6a7cb5b4d6ff5d153147eff579d3e602ed3a0390"}
{"id": "train_071318", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["governance", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "07bd75c93a6e9c11bf0c8cc2f7d863d0712d23d8"}
{"id": "train_071319", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "1f869aeb0629686554eb2b87d0f38d04af6d2f0b"}
{"id": "train_071320", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "d888a79cca5c25aa40dc5bf20d010522e07d1a41"}
{"id": "train_071321", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "82bf81096ee2744bc08b1dfbd802eb58be8e3de3"}
{"id": "train_071322", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "4a52f897e6c3093e4b432ec2450b774f18f4742c"}
{"id": "train_071323", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "cfa9ce83494e1f1871d0755ae2ca3c412fe9bc5e"}
{"id": "train_071324", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "85ea21db941e346494ba415a17ac269dea538e15"}
{"id": "train_071325", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "5a8f8f4ed66ee4b43f08834e541c8b843ef922bc"}
{"id": "train_071326", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "d224da9c218ee788e8f57181c122e65aadf0b388"}
{"id": "train_071327", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "35d00ee9b3f57c3af35d3edc3e76605180b9b265"}
{"id": "train_071328", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "e9cb4d4cd8de6fcc4d5cc2ab5c0eab42e7ae7c1c"}
{"id": "train_071329", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "f4ea7ad165742c25bd6bc17975806331fb231c7e"}
{"id": "train_071330", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "41d045b7e1ff2bf0dfc5c28bd4c64625d1c7cbb1"}
{"id": "train_071331", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "a4b7077f05f22d901a97f9bd9fcf4a583a991dfa"}
{"id": "train_071332", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "bf54c5a88bda7485bf25359740107ce11d308478"}
{"id": "train_071333", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "059b858667e46769576abd14eb25551ff0f9bca3"}
{"id": "train_071334", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "8b217b7e0763372084973dce9259f51ee27cd45d"}
{"id": "train_071335", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "0ce920f85f5a0aedc66330faa2fd7338c01b64a0"}
{"id": "train_071336", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "7384c28528ef216ecceacb17d8d9c160014fd5c2"}
{"id": "train_071337", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "4b790bc69249f441994f9c7d4246f9c23d07dcbf"}
{"id": "train_071338", "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\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f489ca890305b88343358fbbebc8779e73db4a17"}
{"id": "train_071339", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "cc6e3f41dd8b5440e60d965a7111de362666beb4"}
{"id": "train_071340", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "96eb35167e62f8aa4fb0e7e4154abf4eb2aac7c9"}
{"id": "train_071341", "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\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "9d69e434091dda0a4b27efbea148dc2fad5a039d"}
{"id": "train_071342", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "f28086b9a44e20332577ca2f49f2c29823c45381"}
{"id": "train_071343", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "a778921611932b713202311ab0254226840a9a8d"}
{"id": "train_071344", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "728ffb60f5f5697d1f0ef3b98b5f21762cb4b014"}
{"id": "train_071345", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "279ee4df1e7912218898d408861e4dc2f6978134"}
{"id": "train_071346", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "33d5f9408e54afcab24561ab7f3366303e0f03a2"}
{"id": "train_071347", "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\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "9f2b7edd9f3088877f964e48433d0f54ff4eedf9"}
{"id": "train_071348", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "5e129fdfae7424e6d5ec41297cb6d3f655883834"}
{"id": "train_071349", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "fffd6cddab5a191718495f789913c6f88c082a2a"}
{"id": "train_071350", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "a290f24da17447d13890140e7aab503709334fde"}
{"id": "train_071351", "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\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "3c0c8b1928a27bfd629ac494d1f37c1f7476b12c"}
{"id": "train_071352", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "77aaf47b35fff9adf3321ed8625bb8b7c42428ab"}
{"id": "train_071353", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "244d7cb5b2fda6de457112bcb252ebb7e3991155"}
{"id": "train_071354", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "ffbe150a649ec2a4b6b68bc98bb3800f9b6aacf7"}
{"id": "train_071355", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "dc21e450c64a52deb5855b6d5bd7e134ada7d1fd"}
{"id": "train_071356", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "d6b2d5cdcfd44c0adcded8916c59389254b4f409"}
{"id": "train_071357", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "78b8cde330ceda46c906fba92a5def3b1f4fe7af"}
{"id": "train_071358", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "339262e17c493443fc8df1c20e82afa473532e86"}
{"id": "train_071359", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "29473671db3a7b286f0b0bcf9a0e8f490f87adb1"}
{"id": "train_071360", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "75268d49aa3d075476d86c1a1a75f1ac4d8e2a1a"}
{"id": "train_071361", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "8f7f19c7c1d939aab0683addc408ed3d2ee480f5"}
{"id": "train_071362", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "707ad8d0b87ab4069049bfc1611c8646123b62af"}
{"id": "train_071363", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "43c76866fd57f589765a259f77059dc08c205801"}
{"id": "train_071364", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "06aa06577dc7d583de699fd821ee3ca73794b1bd"}
{"id": "train_071365", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "6b7134114e485e27fd66f25d0828e38bd3f68cd8"}
{"id": "train_071366", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "eb4ca501ecb2fe141c8819cd1b325c911985b1be"}
{"id": "train_071367", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "ca712ade3d3ef418ddfeaf398699cb9a4571c358"}
{"id": "train_071368", "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\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "6b118f34f3a85ecebfd0c434743054d00eb56a77"}
{"id": "train_071369", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "5c8d884c5ee78c844a86f9a20d4ed72e9a91dbb6"}
{"id": "train_071370", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "04baacfef37014c4f08bb284c47f517a5570f589"}
{"id": "train_071371", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "ac469529606157d0bad510cd3c963d2c394d820b"}
{"id": "train_071372", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "24bf37dbfc7382f7e2f282b0d37cfa67de442934"}
{"id": "train_071373", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "77c06836e6bed26bf6b3fc3e881b5f2cfccedb01"}
{"id": "train_071374", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "ff4bab78e80c9a9d9bd830443d2b123386cce057"}
{"id": "train_071375", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "775737078edbdab8936bb7f56f1a4caaaeb893b9"}
{"id": "train_071376", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "850769d414199edf89e7bdd3170f7dfb12c22c34"}
{"id": "train_071377", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "cee3b7f2e0bba2e54f65278fc16f3fd0d71ebae8"}
{"id": "train_071378", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "f7b6aabf7ea65262445f8ac7417f90945e37f6da"}
{"id": "train_071379", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "7b39f187e33b9735b9278872f29687c616b35b70"}
{"id": "train_071380", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "3bab1620204b990af755b6bd97cbf42c106bf088"}
{"id": "train_071381", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["security", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "dc2e1e9a71437c14b29274b6963e030578973f52"}
{"id": "train_071382", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "b38a4129ba37a6baa2fa175a8becc5512baacc50"}
{"id": "train_071383", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "63f7e45f705868ef7625cc6c25c4620ff74a8e18"}
{"id": "train_071384", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "feebbf7ce8294bc1067d0c092f1b527b5206f7a0"}
{"id": "train_071385", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "eadda9dd521296e496a495863cec2f033154a601"}
{"id": "train_071386", "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\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "0587b85c760fdf59e2a3c1cd1fb9eb61a18ed0b3"}
{"id": "train_071387", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "db53ef86558bc20a4fcbd8cc4f898e67e915512e"}
{"id": "train_071388", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "dd7bdc96630aafbdd247d6728af13c9f346ce9df"}
{"id": "train_071389", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a7c05b303ffc44f003767010f25119c5de56be29"}
{"id": "train_071390", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "2e7616c1a224e78b788fa392c9611603eff3eae2"}
{"id": "train_071391", "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\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "75003ff159127c1bdc4d0a0df221dd2a40d7c806"}
{"id": "train_071392", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "5741bf6864159eaab234368aeb22b5c20dd9280a"}
{"id": "train_071393", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "e9c61b5e62d470154670488108ed77e83e8d4e71"}
{"id": "train_071394", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f032e42b800f15b1d45d0331bdbf6fd8054172a1"}
{"id": "train_071395", "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\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "fee5f95f6cbac1134747a93509e0e3c326707e4e"}
{"id": "train_071396", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "8fdaf1e1cde9d491f355560c8924203526adcb06"}
{"id": "train_071397", "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\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "eef57603c76360572a33f239328d13d3fae5b361"}
{"id": "train_071398", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "fc79191c16af4b65e49d1db5a8cc23df8f12af64"}
{"id": "train_071399", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "02e87b48342f054281fb46438310557c093e0206"}
{"id": "train_071400", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f2a1c1c7da30344dec20ee847800e2d727e041d6"}
{"id": "train_071401", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "6dfbc89ce1acd687f4b4a4bb43faea8f4cc15cc4"}
{"id": "train_071402", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "e78650cf94093b5773041be25e4bcf65a52731a3"}
{"id": "train_071403", "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\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "ec869f64b3f6569c25c92937b53f85e6087f444a"}
{"id": "train_071404", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "8a62e2b91b9ec02a1e5258c22a935b952d899458"}
{"id": "train_071405", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["research", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "5f639752b77223f09d2e41780ceb8889ad99bd60"}
{"id": "train_071406", "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\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "7e47cbc0fd8eae744867e30f70d0c40fa3558990"}
{"id": "train_071407", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e6850e89a2cfdca98a009d2205bf583ceb6140ab"}
{"id": "train_071408", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "cf22b3555b3003bfc0f0bab94864e834c34eb1d7"}
{"id": "train_071409", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "6aee7ac37bcfb051f0d4654f5ced3e36db14709a"}
{"id": "train_071410", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "a8ceac23746c9cbcd6b8133a880b64561ae8479f"}
{"id": "train_071411", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "b2d1f307a0709e33d583c2ccc3d6ed2830d74786"}
{"id": "train_071412", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4640267f2eb60d22022ed7232af39a200054d8ed"}
{"id": "train_071413", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "6b077c53eb58ca22de3382578910911d195a3263"}
{"id": "train_071414", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "8f7f19c7c1d939aab0683addc408ed3d2ee480f5"}
{"id": "train_071415", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "5255d3b274e572df8d0145e70794b30e299bb57e"}
{"id": "train_071416", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "8c688ab5d7b5e6859a607a34c20c2dd459cf9db7"}
{"id": "train_071417", "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\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "7e166c4638dcd96c49a4078e5a27c954080aa44e"}
{"id": "train_071418", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["security", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "3427ea83fb62ae1937c012cacd8be27718db2b1e"}
{"id": "train_071419", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "205e407254a8a5975357f089d04b90e6156d9ff1"}
{"id": "train_071420", "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\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "2d0f4ee05fabed0d88bfe76f3482933ef6b83f0b"}
{"id": "train_071421", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "787b2b00e935d908bcca36a4fa3179bc9980f592"}
{"id": "train_071422", "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\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "46029fcc7d2f9902a9e5f2dbd01efeb1ab7dd118"}
{"id": "train_071423", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "998ad25cb23c5117a15befbc7a3b13b5d2be3ade"}
{"id": "train_071424", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "7081f9b5811b79cd1ba6853ad375b54d3f43f0c4"}
{"id": "train_071425", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "20bdee81c950ea05742e67feee68c0c00ef166d9"}
{"id": "train_071426", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "1974894efa2ce95e5f91c7dfb020e65cea5fc2c8"}
{"id": "train_071427", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "608de79249dcb385ccfe203ee4bd2197be646b4d"}
{"id": "train_071428", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "624ca98d01caf4847eb0831833d125b33fac6385"}
{"id": "train_071429", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "feca3e010da599c6112c4f2838b48be2d1ef48cb"}
{"id": "train_071430", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "159b3d66bf0c0ec518f79743ccb58c8af4c10a91"}
{"id": "train_071431", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "2f950a42dfd317e1f4c570273bb9a6d3bffb619c"}
{"id": "train_071432", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "790c50f62c352322f2bdfe81908a1660d1c16bf7"}
{"id": "train_071433", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "9a6f7b2aa52e45f656eb3185b57f0350c9056afc"}
{"id": "train_071434", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "f5d0fd588dcfd2bcab65e4b8c474e419d21d0242"}
{"id": "train_071435", "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\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "d20b375672691e6f4f85e1fcb068f8823206815b"}
{"id": "train_071436", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "32c15d839362b4c15a93f7fc8658e1fbf6541f7e"}
{"id": "train_071437", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "3ac6b3f17e6eb65309665bee6fcac11ce7dfb586"}
{"id": "train_071438", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "cb140c9e2552b2f17d7d44ff86d4cf2e1165d563"}
{"id": "train_071439", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "2a3a05c6bec3e0562d38632a82ced67dde72fc42"}
{"id": "train_071440", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "c593107b3db5fe3a94776f77f2d7c4eb48fce393"}
{"id": "train_071441", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "23e0ee7d90cd7583767f398ec5ac5b8c3871bb53"}
{"id": "train_071442", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "2a15299bcadb1902238b96571ad684c86efec071"}
{"id": "train_071443", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "737b25f41e45e68d32a5659c05273a0facc5f42b"}
{"id": "train_071444", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "7f0d3a21c6672819a6bbc888c7793c141430bce4"}
{"id": "train_071445", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "8fe50e6ca1d18ab8e2e37e6a62222157ce278e20"}
{"id": "train_071446", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "b22e00424fa1dd4a49ed2e76ca3c883b930415bc"}
{"id": "train_071447", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "4640267f2eb60d22022ed7232af39a200054d8ed"}
{"id": "train_071448", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "3e14721dcfd915628a81f972b399eb8da79524ab"}
{"id": "train_071449", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "7879a1ceaca7d6799e49ce3ae7fa90caf0e49148"}
{"id": "train_071450", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "413270efac9c51829c7fe1217d5fe61bf1083644"}
{"id": "train_071451", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "f597b21f4124c3354007c102eb4b45d37c65057d"}
{"id": "train_071452", "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\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "cc7e5d5f30b32bafec946cf0c6057b007a844519"}
{"id": "train_071453", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "5850d877e97adc2f71c3aa26fcb5c6e0430c33ca"}
{"id": "train_071454", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "0e0ea7c7037a9b572fab5951cbed0bea77ac9b52"}
{"id": "train_071455", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "9aa8e92de14b890db38dd8d940f39773b3044a94"}
{"id": "train_071456", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "35bcacbf731811ec6ad7e827ca7ef12279af200f"}
{"id": "train_071457", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "dcf0c03b821b56bbe03c82349805686f2114dda2"}
{"id": "train_071458", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "534688190abb99d54e6e3fb13d21be4ae7c47424"}
{"id": "train_071459", "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\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "0f5336922b7430d06ab0c1106cefd2ad2923bff2"}
{"id": "train_071460", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "d04bc5e9999eaa5d560dda9e993fbe305bc3c485"}
{"id": "train_071461", "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\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "245d0ff7c5ee8444d6e01fc6eaf3dec257f7536c"}
{"id": "train_071462", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "8b523e92c2ba6d0dfab367f5b7e8364660703485"}
{"id": "train_071463", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["security", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "90e564ce9b818b2227e6f32515eeb54f9a0e782a"}
{"id": "train_071464", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "5c8d884c5ee78c844a86f9a20d4ed72e9a91dbb6"}
{"id": "train_071465", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c1f02a4b66db37251a489a26173e3b4c41b1ac63"}
{"id": "train_071466", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "cd48e369517186753fc02525e2449866abc628a9"}
{"id": "train_071467", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "2a23d3003ba37c30d60ee3936bce435573cf7e70"}
{"id": "train_071468", "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\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "7c69ed5c858fa7f0e6b739f1b7f48d05db1604d8"}
{"id": "train_071469", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "ccc8c9ad48ad39e1a57dc25cd12ecd3d121b5d7d"}
{"id": "train_071470", "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\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "5ae96faf8e56d9a77d1baea61d14f6bfafd69406"}
{"id": "train_071471", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "7746ef6f2e91735e4552a66bcd18c08671d371a1"}
{"id": "train_071472", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "36adf1f8b1899bfc4e49d6d9f26aed8f58a06e16"}
{"id": "train_071473", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["governance", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "e83e324e95c726a9ef5d48854ddeb740c4ea7c88"}
{"id": "train_071474", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "ae2c42f0ce7dfabb1412177c59324c60713ff580"}
{"id": "train_071475", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "290da7e2caadd9c61617c64348a3d0b280715e32"}
{"id": "train_071476", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "4bf4feed70744ca0670757a48629a4c56a7cf7c3"}
{"id": "train_071477", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "e67e7493ff9ac916ac3f096cb44d3a0a70b20d28"}
{"id": "train_071478", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "302ce8822af15052766544a1983cf027d68db59b"}
{"id": "train_071479", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "2f1c2ceaa2a6d2187236300b4059e8f3dfd6af27"}
{"id": "train_071480", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "b11f7a1b869f5dc92ee01e908abe6619a9b9dd8d"}
{"id": "train_071481", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "8785c4edb3e994abcf85dfda967f4df8ae83368b"}
{"id": "train_071482", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "07fe7cdab22f245729bf30c458310ab5556833da"}
{"id": "train_071483", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "dc29ef4e6ab311b1812cd064024aea43447b96d2"}
{"id": "train_071484", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "016ab9ffafecbab94afa1f2f12049015e179b480"}
{"id": "train_071485", "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\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "dcb35585ec497b593712574ac468a61fac8e19d0"}
{"id": "train_071486", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "8e2b17f448d2763a048d5893b564487496c1c96e"}
{"id": "train_071487", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["research", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "7605bb523ad2ddd1510acb86e180e42ec69e949e"}
{"id": "train_071488", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "bff68e182b3562013e3af90ba90dbf239c044e65"}
{"id": "train_071489", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "d8cbd5a35bbf0f831b928ba3459654772561608d"}
{"id": "train_071490", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a8d8236315cc49b4cf29765c4664b7526aec1528"}
{"id": "train_071491", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["governance", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "8d8614a9386c24ab519661585c0dc1ef933aebb0"}
{"id": "train_071492", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "be92119acecfa3ece4f6750e0c64ba5c80c35da3"}
{"id": "train_071493", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "31037750d913cf1f2eb372f7c3dc6a5a1c695562"}
{"id": "train_071494", "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\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "e00a97e8adb8c18978049a4e373ac3c62d9f6117"}
{"id": "train_071495", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "7437bdd16d801446c598fcf3d5e033724c253476"}
{"id": "train_071496", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "066f682ee748d718aba963751911e6dee0f7314e"}
{"id": "train_071497", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "3f23d879a492914a5a3ec74e3b262b8be17fe652"}
{"id": "train_071498", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "bddaab2c7d8e6be63bdbd16d2eee468ad1e72954"}
{"id": "train_071499", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "a595aab6d7477150b4d73e09e7daf007dec595a7"}
{"id": "train_071500", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "21ba9a0ac717e06a7fceef3ab1a202d926ec137c"}
{"id": "train_071501", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "9724390f08935501ca487aea888ea8f0da78de4c"}
{"id": "train_071502", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "30c63915e2a431a60049b78cee673e5603b1c33b"}
{"id": "train_071503", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f75c69849d56573e691e7130371f5a415df02533"}
{"id": "train_071504", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "f85ab4bc20f5846254a675a3c00aae0330185b6a"}
{"id": "train_071505", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "1179ffb82e2fca3cce6c47eebd448f33b2a6b6bc"}
{"id": "train_071506", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "b201e6cc81065003e77e0e2f825f35e01f05a2d6"}
{"id": "train_071507", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "90157a27fda551ec3b6a385db2b1be59f609265d"}
{"id": "train_071508", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "2103ae3270b0675329b3f1b6271e7f107ead9099"}
{"id": "train_071509", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "8dbfc8b524d230978986c5aa74db24276676df25"}
{"id": "train_071510", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "54d6460ee50b98250b8d91f4d7c5abf4d4104b70"}
{"id": "train_071511", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "9ae48f99e18b3f1ea5627075effecb2e0cf009e9"}
{"id": "train_071512", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "632b4222bf0421574e4ecb59f9f417a1acdf3647"}
{"id": "train_071513", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "d58b70336f4be172276809b01eb31f7130c2acb9"}
{"id": "train_071514", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "5fbd8fea0efb2ecbf448efa9fe18460951bfb91a"}
{"id": "train_071515", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "bddfb5559431d0e9e2c0eec66804dd06d156a155"}
{"id": "train_071516", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "0b0275abfc308b9df3f2dffb86f58e0a7492fcbb"}
{"id": "train_071517", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "ca4657b9ea66bda02f712529e5474bf83624f2a3"}
{"id": "train_071518", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "aa9bed336083a71cd5a8d79fedbcf5cc8ecd22be"}
{"id": "train_071519", "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\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "a3b5a429817d783bbd8e60c65f4adc450eae838d"}
{"id": "train_071520", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "7a32a3728e5f3bb9cfffa50d2d2e7ba7af81a03d"}
{"id": "train_071521", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "40d36568cd5b823bad35b76eef7d9af7b3b94ea8"}
{"id": "train_071522", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "4f86490865ee27e8754264b095b3e4ff9799f9db"}
{"id": "train_071523", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "83c6dfec8698bc2f820925355ef1ceb3ae702140"}
{"id": "train_071524", "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\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "9d675e65bf253363f186a93d588272a404e37fbb"}
{"id": "train_071525", "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\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "4ed3ddd0013387166535b151a7b8df552b064d50"}
{"id": "train_071526", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["governance", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "70d78986822c7500b779b5d520adf99efda79cdb"}
{"id": "train_071527", "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\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "876689a9eda3f2cc613a5760f069d35307ca2207"}
{"id": "train_071528", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f2bb02baf63497ccf1a715caf6e714d78f05b6a8"}
{"id": "train_071529", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "d7c41b94754a01f0b6dc4752ff4d8b8a6c03ae1a"}
{"id": "train_071530", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "a50bb855145f3f0cb20208b58c4e8699212ea2da"}
{"id": "train_071531", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "774098dab07700da181ab491120e1bb1e5c71139"}
{"id": "train_071532", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "63d448acaebe1f7a8a406914cce79fe541216d79"}
{"id": "train_071533", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["data", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "fb14b8ca586968424a38d90afc74ce5bf74d11e4"}
{"id": "train_071534", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "ffa2ec4f10bf3de1bba763aed946558fe8de462a"}
{"id": "train_071535", "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\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "423f86f82078dabedac61c044f575a138c4b6885"}
{"id": "train_071536", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "ef9cc0cffd531b25ddfb41b2a18b9d9b5acc5c31"}
{"id": "train_071537", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "dd9916a9ce2323d95d4591c0b41d02680ed3d59f"}
{"id": "train_071538", "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\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "a07f1a2f16fd622aed357a8b8e20a81b66f98b4b"}
{"id": "train_071539", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "89905f1b157ba450fcf470108e50d12d40ab0ca5"}
{"id": "train_071540", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "72bf017787bd488ee367910a53914fcde812b5d2"}
{"id": "train_071541", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["coding", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "5d5595d0e6b1edf0b6cacab9d534d876a72b8513"}
{"id": "train_071542", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "627b6aed36ff72007c36172ae199eeeee1d8e7a8"}
{"id": "train_071543", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "eb1fca6fd6f839c644ba250229d23c75d84495df"}
{"id": "train_071544", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "877dc78d2a073d07b52bde906c362568293159fe"}
{"id": "train_071545", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "ccca1c2eba7ab99baa1b40a56d5fb51096769cb6"}
{"id": "train_071546", "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\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "44d49c373d545450c6e0ea43878b05aaff376fc6"}
{"id": "train_071547", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "852bc32fe0a9350928ce6103eeae0cd9d78ce813"}
{"id": "train_071548", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["security", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "e95c9de8c6b6d3c44bf668201e353b3e51122d1d"}
{"id": "train_071549", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "c3805a3fe8d639ea8b2c377fa0842e8bb0b8a4fa"}
{"id": "train_071550", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "e2c51e09fac3500bfd1ba83a0d1eba52be83ec29"}
{"id": "train_071551", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "dea798ae39ec5db79d10b901e60d896c0ff0f174"}
{"id": "train_071552", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "64b49fce0f8dbaddc39083fbd9806268a3a5a782"}
{"id": "train_071553", "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\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "b30d480a1f0e83ac6d94e458c5d7f3dc73a5ff38"}
{"id": "train_071554", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b46bfaebe8c0f5989585a2b8d1772195e68c4b41"}
{"id": "train_071555", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "a9143255dddc132db3c07f16509a47b90a351f94"}
{"id": "train_071556", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "72b9a52abed45d8a5b099188a415cb42a4444c16"}
{"id": "train_071557", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "dab6ba8bb305fac32b23724392b557db203e9fe6"}
{"id": "train_071558", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "24c1a9a818e1aa92fa89b2510138341c6ce10280"}
{"id": "train_071559", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "8c44d7a74dffb6d6f8477dc6badf26ac0372e51c"}
{"id": "train_071560", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "dd5d28810ced38ac1866b864e4f36dcac8e4fa41"}
{"id": "train_071561", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "22980b0fdfd5efe75a15f2a1f88c0cf7ed0eaf94"}
{"id": "train_071562", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "cd8f95bf8042c11d03f238523b7c1623b7d853ff"}
{"id": "train_071563", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a842f5d8b003f24b28203bdef5cbec7ffe0d0b12"}
{"id": "train_071564", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["performance", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "10e55051ec0e3616dcd0928df685cf170e57fdb0"}
{"id": "train_071565", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "f3c51fd446a94c8b84e548e408c80fddb6cce7ab"}
{"id": "train_071566", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "3b7412ef2d3434548edeb5f27fa3819f72586535"}
{"id": "train_071567", "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\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "969d832e43a71ec58e885144ff560db45ee5fafb"}
{"id": "train_071568", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["research", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "9c7703d27857930a8ee4b67e42423e7cc8978395"}
{"id": "train_071569", "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\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "d5580800a54e01086405fbd5728665941105484f"}
{"id": "train_071570", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "3c43b2eaf2d034b2e7b3413a669bd8bfc7236a6e"}
{"id": "train_071571", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "e065cbfad85267a6886687990b4288c6e91171fc"}
{"id": "train_071572", "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\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "afbb44e24c93778d7b5ec566707d614d850a100c"}
{"id": "train_071573", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "6e3c7540f2119046791ed4162c5c289502c8c2e2"}
{"id": "train_071574", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "e578865f54a1ab8857234e8733a3e0190ade8f40"}
{"id": "train_071575", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "43bb2f386c22b05dfdaebf75de65b4d2783489b2"}
{"id": "train_071576", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "5c1f61cf0e63510899984de101720f2d47f69a4c"}
{"id": "train_071577", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "3614213b800472787f3815db6a95a4540b55a641"}
{"id": "train_071578", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "84366419df0161df59a54493c1ab75165a55c743"}
{"id": "train_071579", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "76cdfdde061cd5433574ae029ab71c3b811d7707"}
{"id": "train_071580", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "4f4771d036eaf75a6dbea05b67840e08c6a3a081"}
{"id": "train_071581", "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\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "d1515646625631d4e7f4c82bb3caf09b937f6330"}
{"id": "train_071582", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "ceac9490b66c1623e33cce497d89f9e5661e3ce7"}
{"id": "train_071583", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "63a546822f0353a64856b85e0bef244a44e922d9"}
{"id": "train_071584", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "f8bb47af03eaca38a4ef3ad071b8d017344f2737"}
{"id": "train_071585", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "49ad9149c7b2eb3b16711b4ab536cd4156f28ea5"}
{"id": "train_071586", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "11ccee14367fd671e0f2e5b6a5ec9ec4c68a9d56"}
{"id": "train_071587", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "1e4563d9a316749b8b179dc921c97c4553493e4b"}
{"id": "train_071588", "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\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "633033fdfb76d8e061c685dc92c9e067f9e9c13a"}
{"id": "train_071589", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "d349bbee5229bde5ec15600c33f69723c39fac6b"}
{"id": "train_071590", "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\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "da4f7437625335822438968acefa09e9dfcf6b02"}
{"id": "train_071591", "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\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "d316f83580130b422372e694bbde77620b3152ed"}
{"id": "train_071592", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "22c657ee649b24b27be29ca99b70c62f8a13ba83"}
{"id": "train_071593", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "a36817411fbffe36052b6bb933f6d9f8a227410d"}
{"id": "train_071594", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "f40cc848c54405bb1733129f6ed6944f7e231498"}
{"id": "train_071595", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "4681bb93f966b7c7624c84bdb08b662cbb0564bf"}
{"id": "train_071596", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "fb3c7310365cc4c9f25e00642bc7e4cefcdc89ed"}
{"id": "train_071597", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "45b846bc35f46db503fbceeeff1d35f9ca42cab3"}
{"id": "train_071598", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "1814a4724b63fad159dbcba5c14d14c7b5f2567e"}
{"id": "train_071599", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "c8531e5bb07c478426535a667c6ad512b616210b"}
{"id": "train_071600", "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\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "876689a9eda3f2cc613a5760f069d35307ca2207"}
{"id": "train_071601", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "6479f1d1ce203a42d5695cd41bac2170c518cafb"}
{"id": "train_071602", "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\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "e1ea0a57e58c19ac65cb363150e5b4fcedc06906"}
{"id": "train_071603", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "768b11cdba16487b75ab86f95f6cf2b0e03e726f"}
{"id": "train_071604", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["coding", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "9b41bdd138842e530d443fd6c8b44fa73c526541"}
{"id": "train_071605", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "c47a59d3d317275b03f0cbd92302c15eeaafdcb5"}
{"id": "train_071606", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "7bc6d2a6361168697e80f49f0ba148edbc8e65fb"}
{"id": "train_071607", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "0519fdc859f87e5455039159378d215a95480b1f"}
{"id": "train_071608", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "2d39cae1655f01bb19132fd1bc8bbcd4af6bc81b"}
{"id": "train_071609", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "28bd9185e71d50e2522f87dc27dea62eb591141d"}
{"id": "train_071610", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "aff4052573abfa43b5cc1a0e020216cadb4698a2"}
{"id": "train_071611", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "3b8b7ddeec7032e7aaf592da871ce1d29ea1b037"}
{"id": "train_071612", "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\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "dd85ab6ee7137cb3342793095d1a89e2a9adc72f"}
{"id": "train_071613", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "74ab742ebb6b192db4510309697239f89805a745"}
{"id": "train_071614", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "e10917b7369e13819b0993946f8825adffa96ca9"}
{"id": "train_071615", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "b3e37a7fd8f2840a691664c3ee3283e5be52926e"}
{"id": "train_071616", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "4f321550d2872fc933d3ab0cca410f82e64fbabd"}
{"id": "train_071617", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "65f712ee270b488d213b3f8d0a74e66d1881cf52"}
{"id": "train_071618", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "8e8b6f0846b1ecabb12da717dd710b79516254bd"}
{"id": "train_071619", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "6f6a418467f8013c11939c8cb1dff37d3b2d5cf2"}
{"id": "train_071620", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "1e9ef6a8ffcbfd89656461353886b14a316edf4a"}
{"id": "train_071621", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "1826b134c1eb4a551f47468fed6a4fed59431abe"}
{"id": "train_071622", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "ff51d2b938a7f8d0664e9aef9a9d9eb3c5625f17"}
{"id": "train_071623", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "8c14730b76d7608d83a188259a7ce1e90905dbdb"}
{"id": "train_071624", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "7a44f05fb25f88b1cd133ba450726d79a2a7c80b"}
{"id": "train_071625", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "1b02813710611dc2265d40fa3af447526a089577"}
{"id": "train_071626", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "98a4e9bfcae1c230fa455f3ab5b21ea2e7469fd3"}
{"id": "train_071627", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "3c6a6606f666c76aad9f1fb90b22dfd9561eb677"}
{"id": "train_071628", "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\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "50314f153af2fb2146e6775221f2c1e33871fbf6"}
{"id": "train_071629", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "2e4476e11ff4e78df74de6be019038392170d945"}
{"id": "train_071630", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "d4170dede1699e6215ac40e9f2bdadf7b0010e37"}
{"id": "train_071631", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "85bd455faa6c83df067ca386fb1b0e8b125f57b1"}
{"id": "train_071632", "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\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "3994a9d46030efaf3b94e8651ce09e5c928f8cc3"}
{"id": "train_071633", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "7a40e01ad299946b380c47fde21670d4a9a3363d"}
{"id": "train_071634", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "eaf9d23589b6bfd7af27925ba13882cfc70cfb42"}
{"id": "train_071635", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "3cc49d5e05d7418d2f3dc14c5132f3414a9af47b"}
{"id": "train_071636", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "30c63915e2a431a60049b78cee673e5603b1c33b"}
{"id": "train_071637", "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\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "e4b5f4ba831f22caa16135baa79354ed14693276"}
{"id": "train_071638", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "3488b78ddbf571bc505120682294907e9caebd3b"}
{"id": "train_071639", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "476a0d58375b019bea3778b282196c10fd175025"}
{"id": "train_071640", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "67b0377134816d7f28c3505f53e8263d2b6bcc54"}
{"id": "train_071641", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["performance", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "2836e05c681eabb3942fa3371efc5e5de25dffcb"}
{"id": "train_071642", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "ceee2e4759f574da97fb3f07d5ba92fb6feeb69a"}
{"id": "train_071643", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "eb965413aff84fbe32410faccdbcf56955cde9e8"}
{"id": "train_071644", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "61394271a259bc80384d0355e0c97c0f8ffe7230"}
{"id": "train_071645", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "12cdca7aba02cbac7d47614414c375386212521d"}
{"id": "train_071646", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f7c1f72d1aaf3332785f535ad9ff73e2e64e3e89"}
{"id": "train_071647", "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\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "41151031f3a7e6633d502a08214322bf64bda25c"}
{"id": "train_071648", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "69d6117a3f0567f94a125bf5daf8be78bfa3fbc8"}
{"id": "train_071649", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "060cc9c98899bf6a3f4975958b2858c82be27316"}
{"id": "train_071650", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "e1814be69088ea1fed092e28b16c8b8a7b523924"}
{"id": "train_071651", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "616cdf203fcf7883fd62dc3cea1ca7a9c7936194"}
{"id": "train_071652", "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\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "0cc615f989e17d797951782cab4d936e926fe873"}
{"id": "train_071653", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "ff575954b1fc55b83bbff79628e46f80a7065682"}
{"id": "train_071654", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "f47acbdd550d740ad2af768f8cd98d250ad1cbb9"}
{"id": "train_071655", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "1b7d66d88e4966db25cff24dee632c7156e2e7f8"}
{"id": "train_071656", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "acf87143a5c9b01f57feca51ec4cee6eaece064f"}
{"id": "train_071657", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "315f168898838d7eac203876a9e8df2e5e9fdd4f"}
{"id": "train_071658", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "77c82c735b51ce4138cc9d408186fb3b6e9db9ca"}
{"id": "train_071659", "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\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "2932f612a5a364b603ff87b1be3ba7975d7eb5c7"}
{"id": "train_071660", "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\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "d0f9c29e2258ffb66c74db71129c6810b37cf5a6"}
{"id": "train_071661", "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\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "073165894f90fac274fd41124a23db323389bad3"}
{"id": "train_071662", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "1510a925ae4cd6f87ffdc0bf172fc812db724c89"}
{"id": "train_071663", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "e7ef816766a3e9f1a117ee85c1343ee306539b8a"}
{"id": "train_071664", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "843dbed2f0274440416ef79452031bb94bcb0850"}
{"id": "train_071665", "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\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "c09e34ee3e56c21c4a66061200743b34e6cf3373"}
{"id": "train_071666", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "e1ca50fd7242047e5d8be1918b10219237c7d2dd"}
{"id": "train_071667", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a9fbfd5be32368b7343444e840cb7ac2240f658d"}
{"id": "train_071668", "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\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "79e52883972f69aa1d3df7ee4ae46a4e13c527a7"}
{"id": "train_071669", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "46ccc56a958374dfc44472ff61aa09a32b432a39"}
{"id": "train_071670", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "cebe87c5f1d9c5d765f6eab3e8a711c5ed12e1fd"}
{"id": "train_071671", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "9aae8ff711150f880b3b443ace65f6f09207b0e7"}
{"id": "train_071672", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "8278639263a3a3f9975c295584315cf4a225c6f6"}
{"id": "train_071673", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "b8caec23098c640eeb6667c963a5d2a3db77096b"}
{"id": "train_071674", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "2ba3baa114f1788261f950dca4c271ab47d37a12"}
{"id": "train_071675", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "0238c469819413189826a755bd24eb0b2c984a3c"}
{"id": "train_071676", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "82d3d6f0c9e42b8c94cd31698ced14a6de95e84e"}
{"id": "train_071677", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "51d80072345dbd18ca64ff4e4f96a69541a3a0d8"}
{"id": "train_071678", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "f6c8eb0afb07d949f0c57d24c7a04e4b94c09fdb"}
{"id": "train_071679", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "de898c104c022d1eb247a75d3be50946e15b2d83"}
{"id": "train_071680", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "0afdaaaed6e45f233c5f65c5c159b3e106bdc24d"}
{"id": "train_071681", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["security", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "6e2a1b46e2025d383c0fd95a61165d0bda537237"}
{"id": "train_071682", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "8cade80496aa66e910361e277c78d3f3b0d90ecd"}
{"id": "train_071683", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "d9acc5892c6effa153ce310a876a0eb2dce75cc0"}
{"id": "train_071684", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "1b910540a044bd84045dd149750218c3422a6832"}
{"id": "train_071685", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "a26d367c085e0c10a7cfe4dcf015282567b1ee21"}
{"id": "train_071686", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "445994dfb7a62391e6095cc0928b023e61f20804"}
{"id": "train_071687", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "d69a8ce6aaf9762ac2834ddd2947604c4aa88da4"}
{"id": "train_071688", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "94c0560ec0e97afc334c3bb0801db4fe9b646036"}
{"id": "train_071689", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "affbdd5064197a9dcdc9af3db1f8e8306537a896"}
{"id": "train_071690", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a57326aae59cc870a97b432517de58d32bcb9366"}
{"id": "train_071691", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "945c3a3c3a6c28cb1ed367dce93901022bebff59"}
{"id": "train_071692", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "23b328041ee838622a9379d41a1cca9b4a7a8ccf"}
{"id": "train_071693", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "b11f7a1b869f5dc92ee01e908abe6619a9b9dd8d"}
{"id": "train_071694", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "e5025ef17300c15c80c07da3a07e769c4b1314c6"}
{"id": "train_071695", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "2695b90269979e34fe7d43096edb410a82d70866"}
{"id": "train_071696", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "b1be0c5f917f6042cd8551aa310238634f3f665d"}
{"id": "train_071697", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "0f680fb3107c40842b1965b0559d2e0eb5eb9ddd"}
{"id": "train_071698", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["governance", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "729e36e1df6b4901d9cfca662807541eccf5bb96"}
{"id": "train_071699", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "8c44d7a74dffb6d6f8477dc6badf26ac0372e51c"}
{"id": "train_071700", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "7f92d07158011918f37e8e4a8a250e3463e9a56f"}
{"id": "train_071701", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "e519097dca3e6dfc5038dbb8b0394bff9a15f3d9"}
{"id": "train_071702", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a2fbbce6a4e7413d86337f57aec831f1c5bc80b1"}
{"id": "train_071703", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "9c4f090499e344ef3e40cc4e57ec8eff221e2165"}
{"id": "train_071704", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "2e2755c2b1958b8bb6fb317b73b4de5017ae3a11"}
{"id": "train_071705", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "f5bd782567713fd849e2be4ae487b52b52d5201f"}
{"id": "train_071706", "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\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "200fdc2414a5548087b9365c48c6699f6cdb9ec9"}
{"id": "train_071707", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "a90e7aa5dd3a922550ccbd8d81a8bb19ad8e3024"}
{"id": "train_071708", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f18fad4f73ccbcda4b2764cf44a144a829595a8f"}
{"id": "train_071709", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "17bd03ebfb420417a7488fefa789d213cd66a86c"}
{"id": "train_071710", "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\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "3ace4de40664c7cc8935610ce5311d22607a9968"}
{"id": "train_071711", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "a66a5fc811182911b251aed126f79b1b702b3375"}
{"id": "train_071712", "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\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "35939e5952013d77a18117223b52c54ca54cbde1"}
{"id": "train_071713", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "676c9a8a21528142174f7325a299bcb6f600f005"}
{"id": "train_071714", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e266b00c9bc9d6a0a76b4a3d3a03223a7baaa8f3"}
{"id": "train_071715", "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\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "2ba8e41b4fb33dd25837fd82a8bda7eaf6684a24"}
{"id": "train_071716", "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\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "5ad98cacf7678b3ebdbd18e8c55ea6a5f211e550"}
{"id": "train_071717", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "5bd0fd09421f766a8472b3488dfab6fca822abb3"}
{"id": "train_071718", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "5de118773cc4f7ce4e090d210f760f16801e8d8d"}
{"id": "train_071719", "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\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "fa187b39eb20deb39785fe15808cc50a4d38efec"}
{"id": "train_071720", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "8ed6799c37d457ff8e2c57180be3e1b1d61920d2"}
{"id": "train_071721", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["performance", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "cb1ce73bef491be806cb3a2aea1c59c95e6aa056"}
{"id": "train_071722", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "9aa274526ef46fce38e897aa35f6f260e325491e"}
{"id": "train_071723", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "18142e92270d5d9f84c82975ac68daed151c01a4"}
{"id": "train_071724", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "4b75d965b955051ad79e1dcf8438d5b361f3b002"}
{"id": "train_071725", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "973fcfe3eb04472a63ef49a897748eedddf72213"}
{"id": "train_071726", "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\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "7e46d3934818a3a1e8cf939986f1c5bf7704a0e4"}
{"id": "train_071727", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a0840c9572150c70a1c33a57c8f0e5aaa1d0c237"}
{"id": "train_071728", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "a054ecf33b1329420e6809fce079859eda4fb039"}
{"id": "train_071729", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "4321a5faabe0fb4b801e2b37a669e6011cc5cb73"}
{"id": "train_071730", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "713e5e9e5383d9f72e8245295a4afc00b5dd42af"}
{"id": "train_071731", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "d441ed2714cc59665c1cf4eb3b5fb0a5392bf646"}
{"id": "train_071732", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "02d83c4717d7125032b5f86324e9a07bb97e3159"}
{"id": "train_071733", "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\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "0d3a26f9813cc27a0eec059b1a48ab657cde7335"}
{"id": "train_071734", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "a9c36c70a9bfb7112245cca72ac177207f861988"}
{"id": "train_071735", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "3a338d397e8c8bb6aaa21d557ae731c4cf640ab8"}
{"id": "train_071736", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "f6b224e5e50eda76162f2f9d66af3778a29305e2"}
{"id": "train_071737", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "81b13333a3955fac756e9208c970fe383779f23f"}
{"id": "train_071738", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "965b084038793ba67f14ebdef004a099ad589e0a"}
{"id": "train_071739", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "b1068b7d3df884c9ce8a12254856024c628137f2"}
{"id": "train_071740", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "0343e18203f78148940e12c148ca9f3d4750b100"}
{"id": "train_071741", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "7fa3a8fa476aa00b0fbbc91b26a9d76eccda9a63"}
{"id": "train_071742", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "da8cfb30ad82b3d5d51b2a71c6fcc2daf25d3d46"}
{"id": "train_071743", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "6bef381cf87530b692dcdc20da28f63ec966da4d"}
{"id": "train_071744", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "08f902af149364d9206cad1eb4ba10d1bfa27e08"}
{"id": "train_071745", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "d85f4a292dfc9563a16cfebffb3abf8088e2c2f6"}
{"id": "train_071746", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "16c8676f995e85262729ba425e25676ef279223a"}
{"id": "train_071747", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "7f6acc6c0acb1ba4ee8798d5d39bd9c9967117a4"}
{"id": "train_071748", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "43867e2b470d5683e74ecfd0e66456d338c1ae16"}
{"id": "train_071749", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "be62310aaf30c51ac8cdf3e87f7a766aa2b49cac"}
{"id": "train_071750", "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\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "3994a9d46030efaf3b94e8651ce09e5c928f8cc3"}
{"id": "train_071751", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["agent", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "16fe2128233378535a2c520fb2fcfaf904af4aab"}
{"id": "train_071752", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "8dac6d481ef7ca8bc13532c01f900b30fae1e786"}
{"id": "train_071753", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c590b797140c7f138e407fc96349c866dd4e2449"}
{"id": "train_071754", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "2bb73f30aa8ee1c45e9b73c2f8fce08629db35b3"}
{"id": "train_071755", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "601dec19446cb7cf08209b91eb2bb57e70abf4f7"}
{"id": "train_071756", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "3f0fd682929496792bb7768058d9bb95a53b61ef"}
{"id": "train_071757", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "6cea7d758d2b223dbec508b61c056cb40023f5b6"}
{"id": "train_071758", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "fdece1fcd2d38ce45372d27ae64b3b7101cfe39e"}
{"id": "train_071759", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "3fcae1abe03a6d44fab71330789c650287e35b9d"}
{"id": "train_071760", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "d3e2b46255c27a92e0443ec63df864a84001b97d"}
{"id": "train_071761", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f4ba9b0765ab9e80cb2c502f0b1fbeb58d720f6a"}
{"id": "train_071762", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e201390fd9a1e82fd699be241ff60e95ac4dcbc4"}
{"id": "train_071763", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "ecd516f6472b93c68f0cceee4a86dbde3cfd719b"}
{"id": "train_071764", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "cf5a2cc375594a1bf71436e793095c5c0524f9b8"}
{"id": "train_071765", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "9e17ea2c49f3092c07a970cb1d70db5310921734"}
{"id": "train_071766", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "d2538b0b141d28d70bce2226589742393d1f5bbd"}
{"id": "train_071767", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "e5f5ae33a9e0480560f3bc8f4fd80fcaeb3a7ece"}
{"id": "train_071768", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "b7dfe171b8cd3a4d944fab7ff63fd36c5b37eb82"}
{"id": "train_071769", "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\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "1b94a7850cce5419e1143edf8d609cdeff1a5d1d"}
{"id": "train_071770", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "93b64cbd401bc01b76393243e0ed03a27a212852"}
{"id": "train_071771", "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\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "fddb15027c34b4d0de9300d89c981d37e7d4872c"}
{"id": "train_071772", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "01b16b75482c36e37468d97c217e91c6b9ea0634"}
{"id": "train_071773", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f12ff1ec579bf52d785102cb370a17ca61d32733"}
{"id": "train_071774", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "fa8698f2e37c3a6145dc7834630bd8dc35a7ece9"}
{"id": "train_071775", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "3a8881cd02274169dff9dd4f75ab246991995b2f"}
{"id": "train_071776", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "228df8d949ab26de487db331eeed253c46856180"}
{"id": "train_071777", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "7db9244fd291eec9072e4b36f4cf39752a9742e4"}
{"id": "train_071778", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "0b6f3abae0eaf564928e362c5df95c3842e5ff5d"}
{"id": "train_071779", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "8be4de57c505b41e69e1e7e06bf7c51725d1668d"}
{"id": "train_071780", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "19170cb852b511669c8d932d92896aabafc9e5b9"}
{"id": "train_071781", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "023a8924ab224fd020fd893fb0ee7fa7e1ea2e91"}
{"id": "train_071782", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "877e8e3f36ad9738d9186b3051cfeaeebd55b275"}
{"id": "train_071783", "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\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "be8d9efbd70b16fe834e925ace6f0922c70a3620"}
{"id": "train_071784", "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\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "3d53ab34b98a1e6ab354a81d54fb121b2cdcc978"}
{"id": "train_071785", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "528c224793e08693ac57d646948301f8ef182795"}
{"id": "train_071786", "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\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "99ecaf8be1f905af9f95aaaee0bd970fae9ab02a"}
{"id": "train_071787", "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\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "dd85ab6ee7137cb3342793095d1a89e2a9adc72f"}
{"id": "train_071788", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "defa65891460319bd6ab099a13875ac1b990c554"}
{"id": "train_071789", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "bfd466a85a82d93473a5f176d7fdb1b5145d39ca"}
{"id": "train_071790", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "7a99d1321956bf44557ef5750daccc7dbbb67ee1"}
{"id": "train_071791", "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\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "c32a5345c3b9ce5130b39e1c7e9cdc5b2b7abc48"}
{"id": "train_071792", "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\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "85af969ccbc0ab2a6b0381dc9d7f244f56af1e39"}
{"id": "train_071793", "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\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "a1e0a7a14f896cdc4742c5b2835976aa39f87a16"}
{"id": "train_071794", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "122aeff34979aaca88578640582983c02c4705e0"}
{"id": "train_071795", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "0f7f3624deadc681c06547d970e4adb5e41fcb31"}
{"id": "train_071796", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "eafa52dd1479dd73d483fcf474393b257cc26dc7"}
{"id": "train_071797", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "824ee22dc101327bf2ef05fe52ace6a3c1591af1"}
{"id": "train_071798", "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\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f86bffc93999a5e125facfd376c201d9d9773724"}
{"id": "train_071799", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "caa1155ba3ce5b5b7bce7c66cbafda6cb6d657ec"}
{"id": "train_071800", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "c2497300a135e7254c07b68fb3e44a8a5a713fd0"}
{"id": "train_071801", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "96abe519582ea74467f99bad6175c6f4e07349a1"}
{"id": "train_071802", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "f5b8495fd2023aeb7cfc382d4e9ba8048ee3b153"}
{"id": "train_071803", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "49027c5e14bd07896da42c43d80b60f37be83631"}
{"id": "train_071804", "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\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["coding", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "3a2c6268ff44f260ea36ed9ce8057849dda48f6d"}
{"id": "train_071805", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["research", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "1faf6254e00a4eca538122ecbf563bac31b4d606"}
{"id": "train_071806", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "40a28373bbeaa7c28e4ff65a42f4323c5efaa9b3"}
{"id": "train_071807", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "55da5c8331ff707b3112dd117a4d3bd0d95a52f3"}
{"id": "train_071808", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "d3d9c8aa4394efa55a8bd23b1e1cd377040ac542"}
{"id": "train_071809", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "6dd0ae494109ba87e14b90b8aac3c4130787132d"}
{"id": "train_071810", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "b408ceebc2ced89b454ebf7c4183cb0343f61ddc"}
{"id": "train_071811", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "595d032ec62efd28dcd23ee7fc5446188903462f"}
{"id": "train_071812", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "d2016e65802d922b8a06407c1a1588e29490f4a0"}
{"id": "train_071813", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "103102daf600601567eab6ce31e97c97c23a3b31"}
{"id": "train_071814", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "6ae5e368aa53b64dc4c0f48625bbaae2702ce3fd"}
{"id": "train_071815", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "602b305216cb0024700a32c0ab1bcc85479de2aa"}
{"id": "train_071816", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "400e50eb6d49bfed7173b08682dcd1833bbedb86"}
{"id": "train_071817", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "80c0a9bb891f74dc67d6d4ffbed942280bcbd9c7"}
{"id": "train_071818", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "fbb3e71a47a342bdaeb737afc21d41e1385e242e"}
{"id": "train_071819", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "679241b6aba6adf52e5a3eb13ce3528e11362ce2"}
{"id": "train_071820", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "2a9f37639beb96dcab21c7de7d5e33b52da7cee1"}
{"id": "train_071821", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "27f5869b05117b8657bd7591fbdcd858166f7b18"}
{"id": "train_071822", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "27d79fd5d8095c41bf23b19e0ff4f1324291c3b2"}
{"id": "train_071823", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "b6757a2b71785ef68bff46ee6dd6132a3bfa2a2d"}
{"id": "train_071824", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "9ffb2aae4070fa0a622649448a3bb3d8932f66ae"}
{"id": "train_071825", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "bb3ff826aff1df2f7a54cba4b7b50dd329136596"}
{"id": "train_071826", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "6167afb790e69365077c3a0720c885439f598a29"}
{"id": "train_071827", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "957cbc3e7158e9f6f3f4b1dd35970996d0849831"}
{"id": "train_071828", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "f00920b7aae7180356f3e671fbc2ca4e1f5c3bcc"}
{"id": "train_071829", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "2489cf35f72d118598757d19b1ebb732552dc583"}
{"id": "train_071830", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "eb4ca501ecb2fe141c8819cd1b325c911985b1be"}
{"id": "train_071831", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "b13ae8cb149a5124d6f71676969e8b73c068ec99"}
{"id": "train_071832", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["agent", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "e01a9c2fa42c027bc6b886a1088e7c2344586a78"}
{"id": "train_071833", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "7c0665b741b14092e44f83bf8653ac8365ae33ba"}
{"id": "train_071834", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "ef01cfb40d6c912ddbd0ab3f4a02d94155f48f59"}
{"id": "train_071835", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "a0bba63280841063c47f433dff0f42271ee9beb6"}
{"id": "train_071836", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "5049860b1d7aa1795d0410315822b57adab9cab3"}
{"id": "train_071837", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "97f1bf9df1eafdbb9a4082b53bbfe185004f1027"}
{"id": "train_071838", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "28bd9185e71d50e2522f87dc27dea62eb591141d"}
{"id": "train_071839", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["governance", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f57556a1cd7c53bb5e4b01cee65a57322266fca4"}
{"id": "train_071840", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "44a8afe560771e710e3c5320f279bc1909af61a5"}
{"id": "train_071841", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "c32c25f58839017eb139be970ae46046c9293007"}
{"id": "train_071842", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a3dbc39e7f0b7114f2c97ed72d40ec373be3b2a6"}
{"id": "train_071843", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "73ae9b4fd08cb990c3d96975faa2c026997b6244"}
{"id": "train_071844", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "ee271beb1352fe16c99c59411a23535e03101b4f"}
{"id": "train_071845", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "d638f6c58d0f377d83f0183ee0d0fe81e06cb7ba"}
{"id": "train_071846", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b86a227932de3a28ca5b1149cde1f524d5ff51e7"}
{"id": "train_071847", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "11be48e77d1fcb45dfe4f672f0447260063b5043"}
{"id": "train_071848", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "d6b8d64963bf79601fb27047bd524a33fed79052"}
{"id": "train_071849", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "2feddf3eebe3f97d2b961b6fe8064f006e347960"}
{"id": "train_071850", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "b473ee2b911add6c4d262f69780c5b1231319809"}
{"id": "train_071851", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "ab9084556ff64d5a319c31cbf7b0f12524b2b0f1"}
{"id": "train_071852", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "1b282106e4e37b95e977e8bcd652208b9ad1712c"}
{"id": "train_071853", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "9ecd1df054295bccd329c24b8c57330d9b628120"}
{"id": "train_071854", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "ad60217dc60acfcdf6ea89366fbf93ec99509a65"}
{"id": "train_071855", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "a368f13ae2a8d36d135c6656eb8ae3e3a9ba58a3"}
{"id": "train_071856", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "68c5cf14a52ba1287507403fdd0fc8aeec2fb342"}
{"id": "train_071857", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "f9ff0813cca58c052b1236640ab52d7feebf0a97"}
{"id": "train_071858", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "d395a6dc6b5e0dcf2436776bacc2d367812ca24a"}
{"id": "train_071859", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "39349d4a8c5fb334fe0ca9f56dfe7b9ab8e682a5"}
{"id": "train_071860", "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\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a6030456ad994b98584273302e2cf585dd9eeaff"}
{"id": "train_071861", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "ec579660dfddc67944afd5b646a598e8c94b2a7c"}
{"id": "train_071862", "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\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "c95f258603fd288e57d3613f7d5a247d33810b07"}
{"id": "train_071863", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "c820785cd0e067b91d0c7ced7e479657437ceff7"}
{"id": "train_071864", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "40714721e02432ab6a7e601ad6e9641528a2b27e"}
{"id": "train_071865", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "7f0be8ef84b33aba7dff3ded5b7b0624234b98f3"}
{"id": "train_071866", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "40944a382c72ec09d01fea29ae3aa83ff29b723c"}
{"id": "train_071867", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "acc7c0ebb03ed8e366eb1c416efc8dae65f72318"}
{"id": "train_071868", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "7bc7cd908215dac7b522a7c00a8b4312086c0558"}
{"id": "train_071869", "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\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "5585eb384ace4a9499c4cc51b3bdbfd93e2d5743"}
{"id": "train_071870", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "630b5a9fb7ce180b7db8b3ec82d0d3a8642dd5bf"}
{"id": "train_071871", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c095edccfd191e392c70ec4fe01f75825f2e594c"}
{"id": "train_071872", "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\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "78ba56a43af2f3af61424cf9cc8424cff79c329d"}
{"id": "train_071873", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "4d9899f7dfe82c43ae802740f1e49451cf47e972"}
{"id": "train_071874", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "29af3c423f9e4fb39e47b5fa2619f1415ef177f7"}
{"id": "train_071875", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "e680755142a6f94d5ebc39a9ce5b9484a5b4da20"}
{"id": "train_071876", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "c3e38ce2d187af12da6f74757b6ea1c55ee83f7e"}
{"id": "train_071877", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "c8b07e0cf62f13bf876f6f65be884c2e81957ae7"}
{"id": "train_071878", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "bfa670b9c2978994cb79a13a6d005cd0047c3355"}
{"id": "train_071879", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "fc79d14bb8879e4fdda514ceeb4cae4fe17ae951"}
{"id": "train_071880", "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\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "78cd4e348f9660c8a235f2307374bf2426fd87b0"}
{"id": "train_071881", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "ac037d8944dfeecb1303a835926c24b4a0c23766"}
{"id": "train_071882", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "aca9683b8478ac63da0945e11edc73b07a32ad5e"}
{"id": "train_071883", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "16623fc27ecf54d86629038cd3de82217014e1ea"}
{"id": "train_071884", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "2ddd5fe26dae1fea6956ca53cfacc23042e987c5"}
{"id": "train_071885", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "b917effc9cc21f0a09f5b9c24d91e47b71f98f3f"}
{"id": "train_071886", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "5f924332e195a492fa5f78de69d2451a15e93a58"}
{"id": "train_071887", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "90157a27fda551ec3b6a385db2b1be59f609265d"}
{"id": "train_071888", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "72682fb9057eb9cd6615c465502eeb613810710c"}
{"id": "train_071889", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["security", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "ab71ce726379402850858163755a8c0d3ae0d504"}
{"id": "train_071890", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "87dc246719e1145200f83593bf9b89a8607cf4f2"}
{"id": "train_071891", "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\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "80d2cd8d2eab669084b7dac416de1f42780c475c"}
{"id": "train_071892", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "d2ea4b1c3e0159b2d5fe2e87f3677d8132ce3820"}
{"id": "train_071893", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "dcf0c03b821b56bbe03c82349805686f2114dda2"}
{"id": "train_071894", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "5a67a93a7cbaa28b93e421c56516b9cc69845f7f"}
{"id": "train_071895", "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\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "1f5c587e3966e71aa549da3e337ee75e8ab02b79"}
{"id": "train_071896", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "bed38597bbc7c9f4b57b6f59134142eb974d1612"}
{"id": "train_071897", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "e652aa83187851f50cdd2f3d80431351f10938d4"}
{"id": "train_071898", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "c5ea80959cf44257e02eea857ed3b78c45757e54"}
{"id": "train_071899", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "e3883fac3bb79bf2748baa5e53c720435bc631f4"}
{"id": "train_071900", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "8c8417891056429d3e05b7932880289c72e2012c"}
{"id": "train_071901", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["research", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "fb3450052922af191fc49990d7559b2d103c5b46"}
{"id": "train_071902", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "b6d8f6232b2adb69145efa29843a24ff7dfebc17"}
{"id": "train_071903", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "cd8f0496f945ec8521eb97c93b993915140fd921"}
{"id": "train_071904", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "22f6725a68b7977ca14ca8fad2528264afb6d7e9"}
{"id": "train_071905", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "2f1c2ceaa2a6d2187236300b4059e8f3dfd6af27"}
{"id": "train_071906", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "ac6d2df5a8999810028b69133d6425da68a9cdaa"}
{"id": "train_071907", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "e3883fac3bb79bf2748baa5e53c720435bc631f4"}
{"id": "train_071908", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "8ff7ea749e6e5bccbd6eefa5212daaec7c546fd1"}
{"id": "train_071909", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "a29c86f34f1944f55bee9a8335e31a49de69d2e2"}
{"id": "train_071910", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "c1e823576ef56dd1027c07e8bcd2cf31dc6adf71"}
{"id": "train_071911", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "c69f2d909f72c4fa0c5a988fd1a46626ec7e87c2"}
{"id": "train_071912", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "60216b7d7633194db4caa9cd1a52cd3552e5d9d9"}
{"id": "train_071913", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "0d864ea8aab9b9eaec332aec0c163c655939b4c9"}
{"id": "train_071914", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "00739911e10817bb54722633c95bdc0cf69396ac"}
{"id": "train_071915", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "d423564caff6b29314433df0abf1999b55a4c3e8"}
{"id": "train_071916", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "22ce7ef0ead00525f591d26081b2bab1fe470da7"}
{"id": "train_071917", "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\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "3862ddef5dc8ce80c36ed4d9122dc54c7a02ab1b"}
{"id": "train_071918", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "e9c23641b8a773a4602c83c6597c8d03e47b1cc0"}
{"id": "train_071919", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "19630d7124f7403435bd6b172e8bdb24366383eb"}
{"id": "train_071920", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "ecf7706955c6a71d92744cf74c0d4bb8b7e657d8"}
{"id": "train_071921", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "7a8b5beeb1063f6f2e6e06a9a178ef435676d10d"}
{"id": "train_071922", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "de095e20981ca0b8dc9a85f542c0c206d9387199"}
{"id": "train_071923", "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\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "c81b815f6f769fdc70d3494975076d7dc231daf5"}
{"id": "train_071924", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "5606944b37701f63f25898019bbea278c3fbeff9"}
{"id": "train_071925", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "3e6994ddac16e64c23e9596ebb6d5654278a90bf"}
{"id": "train_071926", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "bec7d74ef22bd62983d9d8a6a3d59891dee36c94"}
{"id": "train_071927", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "c66c1a04a3c0aaf6eb94ed66c28bb6c78582c216"}
{"id": "train_071928", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e2728113495a0441a36b71366c0d1a5e4dfb915b"}
{"id": "train_071929", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "627b96c9fe341cd9edd8565de3ad918ed5dbfbd5"}
{"id": "train_071930", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "7c47594b69982e70516ccae274b887d3bc5c6c19"}
{"id": "train_071931", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "b92896c73e30a12c938854efbdd78c2593ee55df"}
{"id": "train_071932", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "c6feffb0192f86663d18c225a9947452fdd4d5e9"}
{"id": "train_071933", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "e02d44a31bbc9b36b463f571bb25ae39c5127f9f"}
{"id": "train_071934", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "76f0e2f5c2fcf88896b59572aba45bd6e77fde9d"}
{"id": "train_071935", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "f886b1402c6f32d1802f11114c9a97702ff41af4"}
{"id": "train_071936", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "fbb831d1c9d9cece6ddb83923c6b7f690d489708"}
{"id": "train_071937", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "1817141f3c45d93176b76a86913932060cebfa85"}
{"id": "train_071938", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "97d9dca1088e97a7dcc7714de3f3048a0300f0d5"}
{"id": "train_071939", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a15c2c23d33decfc9186ac08e8d3478fe50c463b"}
{"id": "train_071940", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e660f230d924ed66cc0bd072688f5b32a351aaaa"}
{"id": "train_071941", "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\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "787c079f7cfe1dd71d24f0e194d145c6847ac8d8"}
{"id": "train_071942", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "3739955e452c383df60ebb8879a0fbc904ae8500"}
{"id": "train_071943", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "6dafaba564acc884f450757c78069b3d1f398afc"}
{"id": "train_071944", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "7f587224784918a0ebc41b319feef146227dcf23"}
{"id": "train_071945", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "888420f6a746982494ff01503c5b86922bc038c8"}
{"id": "train_071946", "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\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["research", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "45e2693f4746d62b0d2889ca5f92466e0bcb7d23"}
{"id": "train_071947", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "b9ab38b8597e2dd2ec43e1460e72ec6d16de9bad"}
{"id": "train_071948", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a5637f7007eea223d5ee33ac2bfb46757ce0e691"}
{"id": "train_071949", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["agent", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "dd376a5cc19ec8501fc74e4448fd10c8d3cbdfed"}
{"id": "train_071950", "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\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "53a9e95634b976f4f629317af4bc158628ebc366"}
{"id": "train_071951", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "9371d02901d48c94716642e770757462a1f2a7f7"}
{"id": "train_071952", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "a93ccb26b4fd1d8419f301a2b2637989ec2152b5"}
{"id": "train_071953", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "aae1c3c4929a5cd7827434c9ec22d96fbcef884f"}
{"id": "train_071954", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "51bf39ffb751a2e3a2550e94b6a00d6be38f3b17"}
{"id": "train_071955", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "7f2f6da07fcc6103b7535aebb9d2c8a4541681fd"}
{"id": "train_071956", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "1f9a3686135d74fafd5942b02f94cce8bb6f9a94"}
{"id": "train_071957", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "11522dda8aae6e31bdb3ccfbde0797a48e1090a0"}
{"id": "train_071958", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "514439219fec4719b993272c03a10db1bbbaef57"}
{"id": "train_071959", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "e9f0a7461595fa44d4b99a0774eed7273881c75b"}
{"id": "train_071960", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "03f83a6aba0fc6f0d805ac31819f9de0f47f31e5"}
{"id": "train_071961", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "4d9899f7dfe82c43ae802740f1e49451cf47e972"}
{"id": "train_071962", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "b5177fba56ed3e48757730f4c86b6f1e2d461c4f"}
{"id": "train_071963", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "c32e652125af2fa2f594c6552f71b1846f8a7797"}
{"id": "train_071964", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "0ef3447b7ffe75216fbf11240a641fc7b8dd7046"}
{"id": "train_071965", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "6639da6417862da547e1a7e0c5f7f70b252371bb"}
{"id": "train_071966", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "d00a08d6522c8ec92bbbcd10bbe67e1fce9da751"}
{"id": "train_071967", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "d45bb2d5ba4dd3eec842589cdfe0cf7fa9dc7747"}
{"id": "train_071968", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "c0cae3cdedaf5fa1fb1cc0edb9b0f4cadc57c81b"}
{"id": "train_071969", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "2bc3592679ad57b1ed3a63ad063961ff761696b3"}
{"id": "train_071970", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "cd038faf2135e99eb3a0ec40a68ecfebdb8cd28a"}
{"id": "train_071971", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["agent", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "4060c38a4567aab89a6c90ead82341e5e0789138"}
{"id": "train_071972", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["agent", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "94b596e830fef9b18c97bb2c438ec27615b84469"}
{"id": "train_071973", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "cae1ca2bbd91d9df64151a4493f1e76bde28e4bf"}
{"id": "train_071974", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "2e10456a6eedbc957f819e5208e2181647954817"}
{"id": "train_071975", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "45c3505a601c417305a8222c482829d4289e1273"}
{"id": "train_071976", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "68e6866fffc444012eefa7487a86c3a3c282e3a9"}
{"id": "train_071977", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "973cee298900f43bb58d1fc0b1c84aa0fe9f7fd8"}
{"id": "train_071978", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "d3f32e9e0185f9aadf25ac09794f8d9be7e21576"}
{"id": "train_071979", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "f62101a1032528617f0df1ca6acb89da57c48409"}
{"id": "train_071980", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "1f78697c18566e12753469554cb542e747a7e674"}
{"id": "train_071981", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "1249c40fb1c7bf2f180384d8a0ee166696cc4b23"}
{"id": "train_071982", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "21ab7502d52ec404152d0c5bc5af38ebfc4fd41c"}
{"id": "train_071983", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "19ad38db1e1666b4fded635224998dfa5afe999f"}
{"id": "train_071984", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "38a33301da710ee8737a234b82ea34bc3600b2ca"}
{"id": "train_071985", "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\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "98b388590f501fd01f9b9ddabcb54e3688239f50"}
{"id": "train_071986", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "b2ee02064ee55d478f2c6508b31f7cca3e0490ba"}
{"id": "train_071987", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "91256f7969bb765b60284e8b08710de18f114c7a"}
{"id": "train_071988", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "9615f8c2b0274d07d41fc79b72579db23e10a416"}
{"id": "train_071989", "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\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "7b85afb6f163821545a158023bf8f71f96ad06ba"}
{"id": "train_071990", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "3298760228bbe4f94eabb0a212b34992166bfafa"}
{"id": "train_071991", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "d99352e43216292bde3cbe2c8924dad17c98e1d8"}
{"id": "train_071992", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "5b1aa6203dbc635c0330545719cf42c849e29eb9"}
{"id": "train_071993", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "41c4525e5e77d1187d0f817553f8599d2b8a2b71"}
{"id": "train_071994", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "33165263f71935f77359e2fd014c12616db926c0"}
{"id": "train_071995", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "e6e5984cf8c3e7f51766879db27b8493d9b86748"}
{"id": "train_071996", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "5b91297cd84ec7ecd71456ace2c326785f8f8f42"}
{"id": "train_071997", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "eeaca1badf40f367601cb97c0a2f0ae2257a031d"}
{"id": "train_071998", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "62874b5c2e54b250c524151699bc80b907fb16ec"}
{"id": "train_071999", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "1fedc694b2575ddfd1a301a0e94a4bb9555f0026"}
{"id": "train_072000", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "b88e27a3d2e44a7bd7d97e381f438687578dbef7"}
{"id": "train_072001", "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\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "775cf0f0ecfa1860c2db259b6649fd9f415972fd"}
{"id": "train_072002", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "684e76ead1f5aeb592089a4a4d82897e805f0edd"}
{"id": "train_072003", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "eaf9d23589b6bfd7af27925ba13882cfc70cfb42"}
{"id": "train_072004", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "d0c9ac51ebb7cdf849c3a25bba1e66c87563449d"}
{"id": "train_072005", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "26791df9aba91da619d1c358c379cb9aa76ca99f"}
{"id": "train_072006", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "8201bbfe13d0dcebc61beb9010d6459f72a3cd67"}
{"id": "train_072007", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "d93ff3c9cf495e1bfa19c5ca73a633b01912d1db"}
{"id": "train_072008", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "6e46ee81bba69b2481f008661b00c1c1e4675fb6"}
{"id": "train_072009", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "26bc4ced981fed726707a199b524c0fe102c0472"}
{"id": "train_072010", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "0f99de0fb16a3197cbebf0b287b8efda19adca98"}
{"id": "train_072011", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a648d16a22a2c07e24bbc9ecd017f78d989da66d"}
{"id": "train_072012", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["coding", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "c777a18ee69fe9645cbf53f8028c56322968ec30"}
{"id": "train_072013", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "8354627c6b78ecb8167c3ae55c2e6113aa1d5b94"}
{"id": "train_072014", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a57a3d339da7e42a445e7c8984c817c65f3d63f6"}
{"id": "train_072015", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e30cb27e2e427064f82c0b1bc70ed665eea4f225"}
{"id": "train_072016", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "efbf159dfde166d420dbd8f7372b5c70ee661717"}
{"id": "train_072017", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "f5b17356c75787b25f188e09eb793f549f58642b"}
{"id": "train_072018", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "c8568aaf365f3ab5fc7e12c417cb2b6bc1f0f852"}
{"id": "train_072019", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "c1ff2b3e9377f28a98fdbaca5ebe17e24a13491a"}
{"id": "train_072020", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "2bd8b660e74824b8f67a7bb58f85164626a9e816"}
{"id": "train_072021", "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\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["agent", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "5847ac4d08a03e24fc826788d98c81c66d2b8dd1"}
{"id": "train_072022", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "6a544906a747ddb6c7ef249710bbf6ba51d62658"}
{"id": "train_072023", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "84b9dd50382016dba128f526c743f39b669acb2a"}
{"id": "train_072024", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "0007bdcd575c43806e52d8cedb568a7ae32d6ef4"}
{"id": "train_072025", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c63612ad248a7a49fe8187cdf818cc8ddf2492fa"}
{"id": "train_072026", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "bfb58c3ed101633ab0d908c847cb9ff03de019e9"}
{"id": "train_072027", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "7632ba95a314f52efbf5e6680e35733a70da9f9d"}
{"id": "train_072028", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "bde45e34bc87ff0764cb951b2a7897faecff1d28"}
{"id": "train_072029", "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\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "cbc429a544baed70a62a8659fd7beb838f88785c"}
{"id": "train_072030", "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\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "b9150124c9f85120997d2ba649ca7bce904ca03b"}
{"id": "train_072031", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "dececa0824306b8b5d3a826899c07fa014a47b45"}
{"id": "train_072032", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "ad1e0ce67f3e222d111bdc53d2439738066e3209"}
{"id": "train_072033", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "410e5b62c71a500c301db04bd4dbb018a59efd93"}
{"id": "train_072034", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "75331a5dcea7fc4e77f480573e03912328181090"}
{"id": "train_072035", "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\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "bc88129c6909af849770a79492889507c3bde707"}
{"id": "train_072036", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "38fb5234b227b0870ffd2897535a0d39d654c394"}
{"id": "train_072037", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "1814a4724b63fad159dbcba5c14d14c7b5f2567e"}
{"id": "train_072038", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "14493869096ddc2694f928fd815e211756724b04"}
{"id": "train_072039", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "daabcb96fba793772c3387caa63a96f9d69ee71a"}
{"id": "train_072040", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["security", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f2ba9e578d8f0b36f4eedfa0d957431d818cbf92"}
{"id": "train_072041", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "1887bc1dd5a3fc81df895c580d78e43842132201"}
{"id": "train_072042", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "0e6ea439004c9d435e7c9f7dbd01f37bff89859c"}
{"id": "train_072043", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "448cb2144506deb3bc90e519b43fa08ee527a041"}
{"id": "train_072044", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "9cb2754210689751d8cc4de2c8ba4cc3941820c8"}
{"id": "train_072045", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "9fe9641397be92ef9f562b2b9a2a454b48d66c67"}
{"id": "train_072046", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "30e2c0eeb29d2806ff60be6325a3adaf8ddc260e"}
{"id": "train_072047", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "080f48301329aa21f526b499e35eb250359c78f4"}
{"id": "train_072048", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "8d282358bd8ba590089c6eb663565597c36c69c8"}
{"id": "train_072049", "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\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "9b20e869fca595cf7d7b48f5a3145db6590fc423"}
{"id": "train_072050", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "a3542cf38c94297e611bcbe5e7e0b8f0f4712db6"}
{"id": "train_072051", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "f7a9c579b586c56f1fbe640c0d8835a5164f31b5"}
{"id": "train_072052", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "b69df4556637e279e86b74df2a736fc7e7c77e09"}
{"id": "train_072053", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "4f4a7e2cbefb5a3532f36b11397c89b4a5326386"}
{"id": "train_072054", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "65ffbf2208afe9e075075d7b33fc7bb504d85644"}
{"id": "train_072055", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "55216feae6d4692e1208731707f11cfe0b445132"}
{"id": "train_072056", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "fe19ca7af37f3017f12ac87226cad68e1fd046ce"}
{"id": "train_072057", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "c5c36c74fbd39d336e27b719233ecdc1d0da46bf"}
{"id": "train_072058", "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\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "775cf0f0ecfa1860c2db259b6649fd9f415972fd"}
{"id": "train_072059", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "c1a196baad71dcccd6958578f641842a9f6a84c2"}
{"id": "train_072060", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "860c7a53c5be46e2a853c550e6ffcbc986dd8b05"}
{"id": "train_072061", "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\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "90daafdc1bca93b5184fd6ac210abcf3cb91aad5"}
{"id": "train_072062", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "cb10615217464c0c0b9f8ba584b3383f8b958e63"}
{"id": "train_072063", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "04ed35448b6ae1d891842b50f2fe7526980447ea"}
{"id": "train_072064", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "a9dc23a13722d16a53b816c3f734045dc9368b57"}
{"id": "train_072065", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "716162ab83510e379a132b06f713e31b22b29778"}
{"id": "train_072066", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "9b75ec8c153182951dfd21c277b9539dbd76d210"}
{"id": "train_072067", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "c273db0ee1d9763ff3c9bef6e1b2fc7cdb6754ed"}
{"id": "train_072068", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "4accf697455ea24bdba13542cfe6b30cfa552922"}
{"id": "train_072069", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "b84f2083f0ba775eb50b4f0f42c633a030810ea0"}
{"id": "train_072070", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "892413a08cf9768ab3da9c16bf297e33fc353a89"}
{"id": "train_072071", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "787b2b00e935d908bcca36a4fa3179bc9980f592"}
{"id": "train_072072", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "61c133795dbf7349f0ae6c5d57a04830636dce99"}
{"id": "train_072073", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "562818efbf50d4784d4cc9ba19448410c83236fa"}
{"id": "train_072074", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "e0dc4051717517018c1ed7d8d1570d5c79fb1cb8"}
{"id": "train_072075", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "d7ca8155eb72d41d7178783da791f0c96588ad18"}
{"id": "train_072076", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "f28086b9a44e20332577ca2f49f2c29823c45381"}
{"id": "train_072077", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "34201a250b9c4de3b5c2273b3e167c3b1fe63c6c"}
{"id": "train_072078", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "1811457b2ffa80955f88f837928dc4b01eeae3c2"}
{"id": "train_072079", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "0baad7eecd7305a3341fe75cce4b0753af6169be"}
{"id": "train_072080", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "f5997184a50b927851701fd99f4936ae21553d6a"}
{"id": "train_072081", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "f59162e0ec84f91e269b3fde76b14c051af21b4b"}
{"id": "train_072082", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "fe17c3b3f3cd10f4a7871b9ddba4243cd059da76"}
{"id": "train_072083", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "26050a1464861cdb173dde305c2fe0a16684ec72"}
{"id": "train_072084", "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\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["research", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c679649e7506f1c8be66b82abe020d745717cc35"}
{"id": "train_072085", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c0b48f952b5375b36c078ea392d656161e589b59"}
{"id": "train_072086", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "a6d8c52ab4c0bf3fb52ff5c5b17f1ae318efd2ec"}
{"id": "train_072087", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "691528075709cc387c08507dc3f6571dd04efb30"}
{"id": "train_072088", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "01609e77fc966844d9c0f2a848a39a9252b23cbd"}
{"id": "train_072089", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "992e7dd7583924c119d441d6b1680b00e382e5c8"}
{"id": "train_072090", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "17974173bd1bd9dc73eafdf47990521403507703"}
{"id": "train_072091", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f5ba03d301ddb4205a1919a4c6fd19dc1d8fe952"}
{"id": "train_072092", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "869681201b4ea62b04ee6597af660162794623d0"}
{"id": "train_072093", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "5aad104eabd7aff882d76ea4dfe1a7e9ef0ad04e"}
{"id": "train_072094", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "c07f5e519633826c26be310bf59e18440146ffe2"}
{"id": "train_072095", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "48d601f2dd3ab46390cf511a886439ba28d79c3d"}
{"id": "train_072096", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "b1f2f1cd2d75da1863d81342ce0aefa9f0567c7a"}
{"id": "train_072097", "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\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "9f9d9e0248b63441ef5af60f7292ec3842fbb0ed"}
{"id": "train_072098", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "e252a8b8f9ec096b8da5a33c584c4ebf4351705d"}
{"id": "train_072099", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "2df251ffbf602fbe9d12981109eeaefb66c5cda6"}
{"id": "train_072100", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "8971790fdb0324789364a7ad5700c24ad6f1df91"}
{"id": "train_072101", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "b584e75aa374dbbd71e0e69160d8c1e20acde387"}
{"id": "train_072102", "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\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "20e58a6e920c5bca771726dfbd0d31a608b9b186"}
{"id": "train_072103", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "ecfc0ea36965999ad6ef502f3b900a0397cab290"}
{"id": "train_072104", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "048bdb5093ec68a0a72eda0ad63a92b3aaf9077e"}
{"id": "train_072105", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "06e516c4c979b34fba472cd9987e07468df8bfc2"}
{"id": "train_072106", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "51ae8e3de7e07719f80c238111b74e861b24713d"}
{"id": "train_072107", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "11c4278893cfac78edc4eca69ad21425c69be431"}
{"id": "train_072108", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["agent", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "13742db05d632dc9e2597f921c321d5806680da2"}
{"id": "train_072109", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "cb05d564f8b2a58cffb61f1c122c0776c2720946"}
{"id": "train_072110", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "417cef637f4e9c400daa8e6ee7df04bd32f61fbd"}
{"id": "train_072111", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "dc4b01e3610ec2f653d917afdcccc9e5d4d97bbf"}
{"id": "train_072112", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "c974ad536e7908ccdd6b80b80f7ca5a3f7cfaecf"}
{"id": "train_072113", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "a5cfa2038956f6fd1cd00167e53001048dbf54a7"}
{"id": "train_072114", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "b11686bcd95217d5838bfe01912f44d67458f177"}
{"id": "train_072115", "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\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "78c7202ac852362798c14030a0131420f134580a"}
{"id": "train_072116", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a36817411fbffe36052b6bb933f6d9f8a227410d"}
{"id": "train_072117", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "40f34e24cedde2819a3d02828bbe40b2eee387c9"}
{"id": "train_072118", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "d45db4bc6395e3668f4ec585636d829fc184112d"}
{"id": "train_072119", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "aa81f086b7e6342c33e9c5c053ec13b7f7207633"}
{"id": "train_072120", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "7eaae5d83e7f89be478b5dec96c1555783de65b4"}
{"id": "train_072121", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "6596cce51b075b3c7e0e51b364d955261b2c04a5"}
{"id": "train_072122", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "bd36b01f59ba9bfa6884a32b4d399b8b9384cc70"}
{"id": "train_072123", "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\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "7e1850548afc816d874fa48f9668de2decb142c5"}
{"id": "train_072124", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "5b0b84538184c6170515ab576f703a4fa27afc11"}
{"id": "train_072125", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "1a46cc43b02312229b579be7b8271bd9f9d713d4"}
{"id": "train_072126", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "4de64016b693b04816eaf1aed2c86e64aab56b74"}
{"id": "train_072127", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "bdd26f29d7354b633a19089f16e130ae5e3cfc64"}
{"id": "train_072128", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "af800bcc9d57cefc93c725f95d44571b741a9105"}
{"id": "train_072129", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["security", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "0c9ad3a4b30ad8ad1e5daf2c5e8323864e3db14d"}
{"id": "train_072130", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "3fc08f50ee762539b60dfbec6f616e33d7597bb3"}
{"id": "train_072131", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "bfc01ad8819a636e8e4eb160a6861486f4eef9f6"}
{"id": "train_072132", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "848bde2a4b5ba786b14d5e58fdadf3d5d585f9ca"}
{"id": "train_072133", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "38e23e14fadbafe024400dfad7712523de3c6566"}
{"id": "train_072134", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "9fa62296ea4d6c6e8567cb784b206662f5d69624"}
{"id": "train_072135", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "cb8c6c9d3403ad6fa514319d66b856f54216874e"}
{"id": "train_072136", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "3bd4d365b62cfe698081b2aeb42c34226e3bf620"}
{"id": "train_072137", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "b77b5fda82532730fcdd3732f0087351e0f8136d"}
{"id": "train_072138", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "b3842402629c59c952a9cb7853ebb5c1980ba926"}
{"id": "train_072139", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "c0fecc1ed16413ada770b3f00914a6488c128738"}
{"id": "train_072140", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "c73d3f6690c87001afb5b0c01cca4e2f6bd07b10"}
{"id": "train_072141", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "a76fc91303d9f809da1485c509d71bf70c00fa74"}
{"id": "train_072142", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "4d15804f8908e9a837dd31d13ff35c7a0a60a481"}
{"id": "train_072143", "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\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "8a775ba41acc62210734cbdff687cd298efdc467"}
{"id": "train_072144", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "d71deaaa62d71b4144410ea580e5cd5c1c61b90b"}
{"id": "train_072145", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "c57454a4da244067f0c863750e23859003e61124"}
{"id": "train_072146", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "aaca6ef8735993247caf475c8df5fcb1fe434d9d"}
{"id": "train_072147", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "ac781bdc139f201ebdb53da417154ffb1dea4201"}
{"id": "train_072148", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "24166a3650cb3b9cdd16eaaf00b51522032e3f30"}
{"id": "train_072149", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "0e6c8fdc5210089f67daa2604f03bb801cefbd26"}
{"id": "train_072150", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "ab98f7b752e45cf7785ce4f25ba5fab647154cc1"}
{"id": "train_072151", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "c32a9f1e74061a065c9434041589c0e9f869b578"}
{"id": "train_072152", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e9b39cd0c7dedaecd8960273646747197d086c5d"}
{"id": "train_072153", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "6b8295ff26debf7a2d11ff346cab0b5875f86d0b"}
{"id": "train_072154", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "f018518258480b0c96e94387d5850785af2ab9f4"}
{"id": "train_072155", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "b9ffb2636bb4ca2bd880262ddc39c4be1357bea3"}
{"id": "train_072156", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "8607a99257d3bf04bf7268952f4d7081d5a1f51c"}
{"id": "train_072157", "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\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "42e2570230da7d06d0b5713fefa827d15f60d3fe"}
{"id": "train_072158", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "c5c696b108f53ab792b44dde3d94aa0e77383609"}
{"id": "train_072159", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["agent", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "ae05b6413aa7f8330a58699b6d341b2caa622523"}
{"id": "train_072160", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "0dfcfc85180e52322d6cae2ccd98d14d8a08722e"}
{"id": "train_072161", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "65c534be723aa72a7a7cc0f241a12a0e7909ae14"}
{"id": "train_072162", "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\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "2346b99f42bf2cd46400d4f273c7ae7bdce5d5b7"}
{"id": "train_072163", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "00f350156cb1c8508984e870a62a24d1363b90eb"}
{"id": "train_072164", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "fcc25575bdbe6f2cab48d82547764ae0245260ef"}
{"id": "train_072165", "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\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "0f9f96f73ee128c659e24f0e85606c30879b471a"}
{"id": "train_072166", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["coding", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "6718d222a22ad691de275c8687a5feeb126bc9ea"}
{"id": "train_072167", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "a7019b72bf03640e650472c1db1b928b44389be7"}
{"id": "train_072168", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "9b0635e883861c5cda540ecd8941188a0b490c3b"}
{"id": "train_072169", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "045a51c9e9c4aa26850d79903e903dfd2f314c50"}
{"id": "train_072170", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "1939a5e7d8b79a6e7ef694713b94b3096468d3d3"}
{"id": "train_072171", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "5c165a31b0233aef789da89f64b0fdeda6a4e967"}
{"id": "train_072172", "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\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["governance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "c5843a9331afcf7deb2637957c6c079e2c8f1d58"}
{"id": "train_072173", "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\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "752166c09b882500e15712b5fd2e4cb88dc8f0a7"}
{"id": "train_072174", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "24c845e2723af903dced58c482deca67b6b02e8c"}
{"id": "train_072175", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "9b67906591ac381f396242ece7b1f9e7353050f4"}
{"id": "train_072176", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "f1c4f8054cc3d266d90052b91cdd621e6aaba9bd"}
{"id": "train_072177", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "4e97dee31bc7030d3ce0685791f97e23f8572afc"}
{"id": "train_072178", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "60647cf4a29a0d8438b6d9ef47acbf32915fc2a8"}
{"id": "train_072179", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "6c3245eeb9e14d48590c62b5a9b222f38a03b094"}
{"id": "train_072180", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "840e943b126ca894fd9305b78e812ead720074c5"}
{"id": "train_072181", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "8339ddd7043f1f609ea33c506edc5d7311e53c83"}
{"id": "train_072182", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "2e7a64cf2ec31335da1ccd19e0525830adf8180a"}
{"id": "train_072183", "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\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "15edfe48cee6ffe781744c65a04a29009f4b839f"}
{"id": "train_072184", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "9248d23334fec512fa589f45b5c2539c878dcebb"}
{"id": "train_072185", "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\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "a871ca5cb56a66f63fed7525265e5de60ee9b6c9"}
{"id": "train_072186", "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\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "95417d2b6cc948e7393e0c32956ad6df2793c1b8"}
{"id": "train_072187", "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\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "875f5e5d42edb5ed9a8f01442f24836354989e66"}
{"id": "train_072188", "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\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "e69bab067e625a46d4106639e6b2a1a281b0915a"}
{"id": "train_072189", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "4c76c2c55d95195415a88645541b25c23fc283de"}
{"id": "train_072190", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "ca46a3ae1a620460e94d42e8015eede3e93e26ee"}
{"id": "train_072191", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "cf53ed968d2e7f248ebb99dbe3e7a73881cbe505"}
{"id": "train_072192", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "db05e2ea1ada67ba4363502b6e62485e4a45038f"}
{"id": "train_072193", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "faf3a368ec906d01e74752a748f0ff32dd3a961b"}
{"id": "train_072194", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "12fe6ac8b194c61f381f6fbc4153edd3584dafd0"}
{"id": "train_072195", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "57e69637dd0055d576397cdb1beec08c008db4e0"}
{"id": "train_072196", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "d99c092be5d8f78ff99fae2d748a5f98a6a51b6c"}
{"id": "train_072197", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "9de7345bb3d5f4b154eec762a0f4f9bc6b4a3d2b"}
{"id": "train_072198", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["security", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "ab94cbd6cf1ac3c28b339fbdd3055a8a5d0e9e8e"}
{"id": "train_072199", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "62db254e476330b77d45cb502f82cf1b93aaa721"}
{"id": "train_072200", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["security", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e202bc5a0b1f06149b6c0dbba9d45f182ab7962d"}
{"id": "train_072201", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f7fd2adfa9f48f61692fff127b6b8d19a5c3f76f"}
{"id": "train_072202", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "21091051c09e1ea4d80bb2a99f159647b8b8f0a1"}
{"id": "train_072203", "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\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "59aa170e106b077416b118b92a9d59a54314522e"}
{"id": "train_072204", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "28921b49b4c0779d8e8f0abe1c64ae4e27992b0e"}
{"id": "train_072205", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "25bf353a755acbdf53fc85de492aa8104862af69"}
{"id": "train_072206", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a51ca02b7fb10efaed85c40c77a35932282d14c4"}
{"id": "train_072207", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "e6de9bab9dc4e4b704989968d867f2565e765974"}
{"id": "train_072208", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "b9662fd07b748e50621ff1bfd2b3e7c767aa23aa"}
{"id": "train_072209", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "299282ae16ae2dffaa87c4bc34b4080d967fb3c4"}
{"id": "train_072210", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "2ae6a3d2e9b285236b58591d45c310e9acb54879"}
{"id": "train_072211", "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\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "5b29812983df215e3b23469279a80b5b8567f6d8"}
{"id": "train_072212", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "818c8576820706624e82ebf30a6b5ca4632d65d3"}
{"id": "train_072213", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "923bb84357b6034a61b6a8c533d3510caab4517f"}
{"id": "train_072214", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "c32fa8df0c0f583bbe0aa45d3a7016421bc5c2a6"}
{"id": "train_072215", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "87790017ec8db8bd43a32a6c0cab80936a454bb4"}
{"id": "train_072216", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "1f0bd007cb5167db7c30d20852315ba8b33c6141"}
{"id": "train_072217", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "dd598adde0538a0bbfd43332c25f5362eab347b5"}
{"id": "train_072218", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "07553415fa80a7a82e1053504056d6eb1bea7d05"}
{"id": "train_072219", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a5d6a077ae8b7df04ccd68a476cccba728adcc38"}
{"id": "train_072220", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "72691c237e1049238e511f2ebac0e422f359e9bc"}
{"id": "train_072221", "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\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "200386aae20737e8ff6d068bb439df87f6816700"}
{"id": "train_072222", "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\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "b9c61326fe9be13e1a9ba2ec760bd1ec4b216dbc"}
{"id": "train_072223", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "9cc7d9df5bb143c7e83b8965a1eaee36f21f69e7"}
{"id": "train_072224", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "28b6c4aab232933797a96e08995b2e0e6f175766"}
{"id": "train_072225", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "0936adefc23ea299b1bc5732af978aaf223312bf"}
{"id": "train_072226", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "2f8112607725779979219aa92d0983a96dec53f5"}
{"id": "train_072227", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "6b544be6acb2f3c52499f44b7f856f3674f8fd08"}
{"id": "train_072228", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "5508b6c994bc717d804f6c7ad7e3bf9d0fd256b8"}
{"id": "train_072229", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "81508e760d9403d6850dbe4f306e8050f2550b44"}
{"id": "train_072230", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f4cade7d8c8e935fa4cd3712ea8b5a6270c81b58"}
{"id": "train_072231", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "5634594a5900bf7bdebad0884bf05041cd886f48"}
{"id": "train_072232", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "f082ff81a8456cf279ede37563b6f846e0acb549"}
{"id": "train_072233", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "2889cdae22f80fbe919b99157253c84601243ffe"}
{"id": "train_072234", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b514d3b2e98e5e07715f3d2f8bc07ab64bf98a6a"}
{"id": "train_072235", "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\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "52ca68481091fc42a757e0e17d1e07616cc099f1"}
{"id": "train_072236", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e722868d1ec4db682862eccc55812c1b6b315742"}
{"id": "train_072237", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "09ee9381369da08f3351be8a2f743ca853192597"}
{"id": "train_072238", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "de91afe55414cf6a15d2ae5a9f97b6f1580e5c8d"}
{"id": "train_072239", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "63f5fc8e110c2651d97ad904b58c1d07aa32e574"}
{"id": "train_072240", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "67947de1454843c21ebc7437a6d127e6b4848ca3"}
{"id": "train_072241", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "01ab69ba6b084f4a5165c969192a6f8087abbf36"}
{"id": "train_072242", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "48d0f8ba61d01a4b82982b7e34de3f9b139f8743"}
{"id": "train_072243", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "8d3855e332e9b440d3a161e80284ae3665468f3b"}
{"id": "train_072244", "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\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "7869863c1459f8ad8477ded6ef7f1bd7bf8d6fef"}
{"id": "train_072245", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "d0a7fa16acf5fe1ad1615833fb516f142fbf5c43"}
{"id": "train_072246", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "55216feae6d4692e1208731707f11cfe0b445132"}
{"id": "train_072247", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "4112a0fbbd574836c8b549ceef67af59da9bb79d"}
{"id": "train_072248", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "da66f6c91b274a0a491a189f4fe7c03a94b85d40"}
{"id": "train_072249", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "cff9c5e6cf36a0318ffa887000bc60807765b7c6"}
{"id": "train_072250", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "477c1175a48d7d641dc12706bcae5cc6f437249b"}
{"id": "train_072251", "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\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e9dd259f5df1ad71956b7418facb4e1ead214433"}
{"id": "train_072252", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "790bd047af69816a4081b718fcf71ea55ae24621"}
{"id": "train_072253", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "68768083e7ee41079a175c9723263f25631d5626"}
{"id": "train_072254", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "9e21465bcc63ecafb1166a6b5dd65a24d956a02e"}
{"id": "train_072255", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "38a33301da710ee8737a234b82ea34bc3600b2ca"}
{"id": "train_072256", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a7d56e30a0704055f054406dbec94ff2bd7fa051"}
{"id": "train_072257", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "ebff7c5f4a1fc88f698fed256fbc1c44d786cd48"}
{"id": "train_072258", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "2bd73d61f22d4e7041fd6be1569c1e1cf8785a5a"}
{"id": "train_072259", "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\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "169560cd9b0b0ce69f6193d4cc2b5a3098baed0e"}
{"id": "train_072260", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "0ce97936c9601d294cde82e2b405407efce5c112"}
{"id": "train_072261", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["security", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "0f6f260a37e67301165692a565735807eec1af8e"}
{"id": "train_072262", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b7a6439d16998f00e537ad237d0a57e2fc27e8d6"}
{"id": "train_072263", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "c9cc07ab9f264c1463316c1650cb4b0d0742ebeb"}
{"id": "train_072264", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "76f8b6577a36eca868c2edbcf145b3533b5cca61"}
{"id": "train_072265", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "249cfb91c5ec2716db868ee6c2aff11da66dde6f"}
{"id": "train_072266", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "606733dbceb97a78ddfa399687e8f38e0f9ab2e2"}
{"id": "train_072267", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "6e9112ea814e01faf76bd8f5d4aba6b883131701"}
{"id": "train_072268", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "2afd7ce4cd6222679c954e5ee227fd22bf311316"}
{"id": "train_072269", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["coding", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e4b1f098e79f15c58cb530b74d476860d435e0f5"}
{"id": "train_072270", "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\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "906c1c9059d072208f3b5707e5f54067844674cd"}
{"id": "train_072271", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "62db254e476330b77d45cb502f82cf1b93aaa721"}
{"id": "train_072272", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "b8eff0eab40e3dd459706ca240e6912cba93fac7"}
{"id": "train_072273", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "0238c469819413189826a755bd24eb0b2c984a3c"}
{"id": "train_072274", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "12750ed383657e41cf9152038fa34d912d20016a"}
{"id": "train_072275", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "4c1ad3a6bff79287f71c27c9b2b44c8885a7c51b"}
{"id": "train_072276", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "70d64428e7ea7b411a016cfc6f6fa3445f21af77"}
{"id": "train_072277", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "d5e56b4daddcd53415f83f61fbfb6126fcc5e5b8"}
{"id": "train_072278", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "8fd7c6c1e87d97fe7b78b599fa4bf2a7d85101f8"}
{"id": "train_072279", "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\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "43312b87109dabb6789c0dd1bc857a5c0be51340"}
{"id": "train_072280", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "db419701eb9b8819898d52379cbab52e782d18ab"}
{"id": "train_072281", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "5ef3dadf5fb9b8742f4228b3a6af3f97f456f173"}
{"id": "train_072282", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "362f2f9f36070cbad93b6f436e090a456c049660"}
{"id": "train_072283", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "8e029ead29076d49c02c337852c065ae2278325d"}
{"id": "train_072284", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "de3decac0cd0fcc3b90d11798d78292fde539a5f"}
{"id": "train_072285", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "982ea1e2ac47ea46a30011ff6d321ea0fec69934"}
{"id": "train_072286", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "ecdc25582fad570851bee7c498c9c87aaf5ce8e1"}
{"id": "train_072287", "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\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "fb015b3cd8bd8154da4991dcf1910689adca2ff7"}
{"id": "train_072288", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f1af1cff96f9bef9c278709214ddbde817b4bb1d"}
{"id": "train_072289", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "8cd289ccbdb3e0835cad31a674880aabebb1a33e"}
{"id": "train_072290", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "5c6112dfbf7ce8f3ab56c4ff06beb20fd8e414dc"}
{"id": "train_072291", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "0968c03ea38c45e5e27110cf91ed9f1610a22199"}
{"id": "train_072292", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "757d938eb113915c294cfa97594d02747b0b5643"}
{"id": "train_072293", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "bd94b74c4bb3d74a2dbb118ddc91931d5fc20702"}
{"id": "train_072294", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "dffe0bbc840323f93990d3a33ceb53d8759e8a72"}
{"id": "train_072295", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "1ac3f35a57787d323c1e3df25b48f46d1f3686c5"}
{"id": "train_072296", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "ddb3f3b44057b9011c747705afa24418f40671af"}
{"id": "train_072297", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "37f77c4d2da895d11d438231269f962553b013af"}
{"id": "train_072298", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["agent", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "80629acbaf9ae6d87e6007805d88bf91b777daf5"}
{"id": "train_072299", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "4d408843048011b8c4dccadb8f3b0b80f2ef1eae"}
{"id": "train_072300", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "e0407a3abb26065ac1649e4b67efe2702c853e4f"}
{"id": "train_072301", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "2b420a87d6019bb8355c85b5f6f9d10428f731d0"}
{"id": "train_072302", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "c274e2426ee4ae20c594568ca9117374049562ea"}
{"id": "train_072303", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "1b0c647887cd81527035b48427d2ca08949deb3f"}
{"id": "train_072304", "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\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "059a5ce9e7f4eccb31f99df1636a90b87c004961"}
{"id": "train_072305", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "0ec0e424d2b2728e508e76c9c5f12720c40cd150"}
{"id": "train_072306", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "5b54854a426e463d45ed4ad9aef7533872b805c3"}
{"id": "train_072307", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "6204c6fbf15a165326d7d55371698553ab560d5c"}
{"id": "train_072308", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "b7a6439d16998f00e537ad237d0a57e2fc27e8d6"}
{"id": "train_072309", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "177448a18483a1b93b937dbb703ee22ac9a8362b"}
{"id": "train_072310", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "318db128e0b252d9f9b3d1f7ae9202aaff076970"}
{"id": "train_072311", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "094a8412e26f400f1fa42c2f4960319ed4abaedb"}
{"id": "train_072312", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "6903a00bab8ea1c20b55ed9335d77e6dd5a9e7e0"}
{"id": "train_072313", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["governance", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "a2f1c0e5f5654f5f41e3a660e9b12f544108a81c"}
{"id": "train_072314", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "28b6c4aab232933797a96e08995b2e0e6f175766"}
{"id": "train_072315", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "69e41ef4030cb0463b9daaf039990dc3e5c38254"}
{"id": "train_072316", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "54ffc263254b411531806bb75fe5af89881cb11d"}
{"id": "train_072317", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "d9800199d7d415cd72f9114b9717617fd091805d"}
{"id": "train_072318", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "3d408bb18f315ce30725f226f3103e3e7addd271"}
{"id": "train_072319", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "98a7b1aff6fec99795eb438687a470e3c0e11480"}
{"id": "train_072320", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["security", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "1cc71091134498597c0104778ee74c5256f72e86"}
{"id": "train_072321", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e80515b4b5cee1831955753346ddc2b4e56ac0e5"}
{"id": "train_072322", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e3b1bd1eb0389b667db8fded56f25f695fe9cb87"}
{"id": "train_072323", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "1ba3d0f41b6512141802fec8ba97ed0c70974d17"}
{"id": "train_072324", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "1793d73fd443d51589cf6f63c1731efaa37b55bc"}
{"id": "train_072325", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "bd25d1bb807db30e2fde4c8c2e33a98906310155"}
{"id": "train_072326", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "dfa970700fc32b30309086a906c68798ba73a3c6"}
{"id": "train_072327", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "e649938b8e9ac8c2d0266093800028ee617012b1"}
{"id": "train_072328", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["research", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "9893c1c3acc58ce9990f3bd4ab69111bedefaade"}
{"id": "train_072329", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "62f802c5feb772d1368b6ed4fe9e3c7f17cfac9f"}
{"id": "train_072330", "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\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "884bf890f028ac74011d2cbc3623cc7962b11fb1"}
{"id": "train_072331", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "182655eeaecf38f431b6db425568f46031e5a2d0"}
{"id": "train_072332", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "fafbe48c23e6e9b7b8be0c5fee19783c92fc7a77"}
{"id": "train_072333", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "fa504d982b7d6f52785266e48348681e8c4988be"}
{"id": "train_072334", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "1029e6b7b6015a72113d3e552368621f23a48c22"}
{"id": "train_072335", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "730a9a055557999f12d38621d506dd69dc95f7c8"}
{"id": "train_072336", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "13b10e706d7cf5a7c6b9ab3956abc28ec34f5e7c"}
{"id": "train_072337", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "3e4c60f54461b95484b8b49a0380e2b18d69ef06"}
{"id": "train_072338", "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\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "2fc1d81b8bdedf0fc9d5cbd0bd5c412d7adff058"}
{"id": "train_072339", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "77b473a89daeaf0715f6262cd12af5cd3aefba70"}
{"id": "train_072340", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "ba994e59c5b48d1e850bef380c026db5d172c46a"}
{"id": "train_072341", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "48a5c18055426289343a5ecf90c7174a42f88554"}
{"id": "train_072342", "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\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "72f5c7ca6bef009c26dea177a36fda5eb2ad2dbb"}
{"id": "train_072343", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "6d0f6f8604596d58fc1658f7df5839e93673a425"}
{"id": "train_072344", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a613372c9ee58cc07d7afd55b496357d8b2a437e"}
{"id": "train_072345", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "95f41c039794c2007a031a3972efeef39b873fe1"}
{"id": "train_072346", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "11a695c56b44af449e593e0c9f00f234a3fe69b6"}
{"id": "train_072347", "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\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "84d056b31a865a69e894ad8face63721e93f2313"}
{"id": "train_072348", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "79614b7e60fbdca2a6234769b2e843609e20bbc6"}
{"id": "train_072349", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "a5b1aae5d7f5e00cabb0f92f91a1ab2b598155e7"}
{"id": "train_072350", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "1b1bf808c8590f6dedfa6f43cdcabfe0792abc68"}
{"id": "train_072351", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "25f814fd3f75629e613313c9ec608efe9ff14bb3"}
{"id": "train_072352", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["data", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "79158dfdac0a615aa893975f34528b115dea9625"}
{"id": "train_072353", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "c8ee561fcc9fe7147a84910b1be09941fc118485"}
{"id": "train_072354", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "30e80b8c22baf43dda00d0cb2f54ec01d904e04e"}
{"id": "train_072355", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "19ee3ab6ba73475970f15cfb4706ce9d144e2862"}
{"id": "train_072356", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "0830035c8ddaa7a3ee5baafaebbd49bcfdee2166"}
{"id": "train_072357", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "ac98339a542c716d97f1670d5b66839446aad92f"}
{"id": "train_072358", "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\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "ede3a8bfd50fd9e889090e9a54969afd1ad68d63"}
{"id": "train_072359", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "e46449e4a4df9b7c405455f95c6e1afa4b1461de"}
{"id": "train_072360", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "a373e65b80acb695a12ae8427e568bd3c7280f3d"}
{"id": "train_072361", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "f6423adcd70bb7f5406407f3e9e6cee39f3b929c"}
{"id": "train_072362", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["research", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "7e97f658799a72a35b3c60de1a47b3de95034df7"}
{"id": "train_072363", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "750c3caa5b3cc72720965a3c657436249cc3a96c"}
{"id": "train_072364", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "c9d7a4af961c853f6117a86144e8dea86e1c0360"}
{"id": "train_072365", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "b969b5cb3b9df5b3c27d654748930598e48bf1d7"}
{"id": "train_072366", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "58018a8ddc9d2e82511ead7445832967dea1e4fb"}
{"id": "train_072367", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "10d9773fbf00f99974991f060bccecece925bb38"}
{"id": "train_072368", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "d34566264bd00bc4f99d9b559fbd311c33256892"}
{"id": "train_072369", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "7a06db6fa5b6fb5462b49fd1516e9019d7b6046f"}
{"id": "train_072370", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "432cce4b1e9027e015c129c11e108584a0986189"}
{"id": "train_072371", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "55ba7cb1bbf697f52807962c0df8ad0fae1d6002"}
{"id": "train_072372", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "61817741a4d279fca2db42c4da104c72bb7008d2"}
{"id": "train_072373", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "9abcb17007c37d903fecfe817a50d8376deff579"}
{"id": "train_072374", "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\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "dcfa9c92bfa9be6e92b4c1c9d5c0d78095457502"}
{"id": "train_072375", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "ab98f7b752e45cf7785ce4f25ba5fab647154cc1"}
{"id": "train_072376", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "eaf09b41539299b2db2d8353f22670b6fd5af487"}
{"id": "train_072377", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "56868fb8e7a21d2c4e7d33900e93a5cce86ec8b2"}
{"id": "train_072378", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "b8da598182720a30e1281f99df93d0f1f85fb0a4"}
{"id": "train_072379", "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\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "e52929bcebaf5dc7be1ba0ec2e1e4c16cb127eca"}
{"id": "train_072380", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "80df5fcf5eec594b76ebb2a2bba542f0c1fe8191"}
{"id": "train_072381", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "d805d831cd9bee0f54c29f93e46d8f457147ee1d"}
{"id": "train_072382", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "3298760228bbe4f94eabb0a212b34992166bfafa"}
{"id": "train_072383", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "f80c0e7ab238aa7f8f3ef8f372126daae3360d8c"}
{"id": "train_072384", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "2ed4ca1dcf9b4b7b07c908505b47d46d5008a6fa"}
{"id": "train_072385", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "1e9689e3436da2fd8dd6df896790b153e1544c89"}
{"id": "train_072386", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "0fe8090faaf34e588a29f7106ad0f6b83426b96b"}
{"id": "train_072387", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "15f35bb90aae13f2d8feb173245456b8ad74e5b7"}
{"id": "train_072388", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "6c88bcbb846ba50779200231ae98f1b41d6ff450"}
{"id": "train_072389", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["evaluation", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "01f0ede8641d7d45d31b57ad26694d5fea03b732"}
{"id": "train_072390", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "b461d1c2102138be01fa9dcd97204858743ca8c5"}
{"id": "train_072391", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "41fdd2f570ec1a06817099ad429e6a8d1ee9fa13"}
{"id": "train_072392", "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\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "8c44c3c65b753fb0eca24f7d6fbae0d8b66cd54d"}
{"id": "train_072393", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "974c1db6eeb2e9386dee94b780ae18837af1d69e"}
{"id": "train_072394", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "7c360b73844c9cb3841e3e2683d51f5c5bd81c00"}
{"id": "train_072395", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "27f1f6aed57a5236be66faf440a5f7f48d4e7ffa"}
{"id": "train_072396", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "768b11cdba16487b75ab86f95f6cf2b0e03e726f"}
{"id": "train_072397", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "263f74dbf129d00724ca89a9b40c13d6be10fd8a"}
{"id": "train_072398", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "f2bb02baf63497ccf1a715caf6e714d78f05b6a8"}
{"id": "train_072399", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "f007f2995b7ceac2f73016122d09f279ab015d51"}
{"id": "train_072400", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "5160a31db57211ea02af3e7c1f0249a1b2c3db1c"}
{"id": "train_072401", "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\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "b8dd95467c58f8093929d87c431bd67abf4ec173"}
{"id": "train_072402", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "10efeb6562ac55455023e67150b49013198f5636"}
{"id": "train_072403", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "36429dff918d7ba6ece893dee39e2a0d0aff06a2"}
{"id": "train_072404", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "b0647451eeb3b4dc85d02ad687a3544682fa5d66"}
{"id": "train_072405", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "311cb65db609b52eb74728912a21afac3df70806"}
{"id": "train_072406", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "c52401ade327204540f661e56068d731d88b87f6"}
{"id": "train_072407", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["performance", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "189e1489df6b93c0b84a5366e439ea734d4e67eb"}
{"id": "train_072408", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["performance", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "0755758dc5b8a556dea54e246fea74440fa34176"}
{"id": "train_072409", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "35bcacbf731811ec6ad7e827ca7ef12279af200f"}
{"id": "train_072410", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "2ab872a89197197e0df2ae3561efc634d3a15721"}
{"id": "train_072411", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "3351845424ef58085749162768503f51829f7582"}
{"id": "train_072412", "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\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "84d056b31a865a69e894ad8face63721e93f2313"}
{"id": "train_072413", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "2770793fec002f3cce4e3c3bd5093ce4f8f7c351"}
{"id": "train_072414", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "ae4c1a26d4676010b3484ed9b5c92b9ab33bdfc3"}
{"id": "train_072415", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4ff37dac9b9c1c966a9a959cc41676fce61158b9"}
{"id": "train_072416", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "bead31dad8c126dbe8fdc4329b101e281065b3f8"}
{"id": "train_072417", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["agent", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "9a5a6a0e48d64b9a50f7fd80ac05f06abd5637ce"}
{"id": "train_072418", "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\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "d355335e83eadf5177952573e5a56bf81f7eecf2"}
{"id": "train_072419", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a08c88b646547062ab5ab82686403d258a655d61"}
{"id": "train_072420", "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\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "5e568295c03f17cefe5845316120b69868990146"}
{"id": "train_072421", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "ae6319b29c3d21b0d9e25048fded357f17e24b48"}
{"id": "train_072422", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "023b5eea301dbcb2b195fdf4d1ce58759bf612e6"}
{"id": "train_072423", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "0b39f5be634b09454f3910634d877b1a21bb6b80"}
{"id": "train_072424", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "7e3d6ac9c4f7b2c729aca04805cbb5b5e700068c"}
{"id": "train_072425", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["data", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "75b55a5fa591b1ef59b2f6a3db6982bbd79459aa"}
{"id": "train_072426", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "b320f1a9bbbb0607c1561cdc3dbd5283ceae93f9"}
{"id": "train_072427", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "9b6077b0d6598886e25650b81986d4b437cfc5aa"}
{"id": "train_072428", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["agent", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "20be0bf65e9751464141dad2af68eed21d51d068"}
{"id": "train_072429", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "d4005cd141c58ae884571ede05bb039e70b6c2de"}
{"id": "train_072430", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "d84c11e6e4217f29ef7cf5c13115b7378d6be221"}
{"id": "train_072431", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "9d1bd38b0228dc20d74698908f026b35efe5bcbf"}
{"id": "train_072432", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "760040f677e44e8a6f86d380bd46bb4f336e4601"}
{"id": "train_072433", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "86c66d35a2115ae3410eb134a7e58e4b0cd4d654"}
{"id": "train_072434", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "70040b3805f683f190a59e211b20bf208174d544"}
{"id": "train_072435", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "46e99d60a6d20bd254819bb3146f43cfe6e02962"}
{"id": "train_072436", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b35a2dcc23055919149a00002f8e7995f8fdce92"}
{"id": "train_072437", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "f3002e52bc2fb27e12f9f8dfb5d3dc9d215fdc0f"}
{"id": "train_072438", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "4260092749a7eede393c6b9fe841b73289d962bb"}
{"id": "train_072439", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "dc4bdfd6e750abde08ce8cb481feea5178a5992d"}
{"id": "train_072440", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "828a8448b4041e2f4fc7532213f90d5f9e18c585"}
{"id": "train_072441", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "39dfe2411dd153cfc8b89ceea439d6c7e793df30"}
{"id": "train_072442", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "052027b9168f83005ef5a76cfdfcec76c5f894bf"}
{"id": "train_072443", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "61a185156504d158778ad677892b0dd4fbb96efb"}
{"id": "train_072444", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e2b67df2b04b911826517f0a6b6f1e2049284516"}
{"id": "train_072445", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "fc74af4c92d2f248306a2ae1b951fe2104ac1498"}
{"id": "train_072446", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "4e3ffe14927d1e4779e1b5f5101ff27780ae1f3a"}
{"id": "train_072447", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "62136a0f4596ac4d56c434507d2d64a206a0deea"}
{"id": "train_072448", "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\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "7bd72a2d88e96d34748f85146836cbe0aabc8338"}
{"id": "train_072449", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "e798bb5cab8fc8982cd4e49d5ec8bf0cd51b4e56"}
{"id": "train_072450", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "bbe8b007a8337940404bef5cd5e574040a64b49c"}
{"id": "train_072451", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "d2f774e8c3d69caf9003e44e3be0fc9d11bd91bd"}
{"id": "train_072452", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f0e223698f55a3d9c8d29a0c6d98d6ad52371f0a"}
{"id": "train_072453", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "93b0c10c5716b1c790603e8741414e9be56cba6c"}
{"id": "train_072454", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "9b34dba91b148c785f3704155a9b00ae484fac9a"}
{"id": "train_072455", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "ecadab1cf90b696fb283bcd65c8532e6f509dbbc"}
{"id": "train_072456", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "269167c2b4f6462dc0182ceb4e179a36ba50fafb"}
{"id": "train_072457", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "f12d009fd94e67e8eb26e6d0fb9b3b33a2cdb41b"}
{"id": "train_072458", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "be19a0ba8659bb34b6c7d0f2a89241eced244f66"}
{"id": "train_072459", "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\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "2ad406382a090ae31a6591f49240b0ea2616f604"}
{"id": "train_072460", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "e47c43583d4705ca8ab296763983f6b04701fa6e"}
{"id": "train_072461", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "8d45ee23d46bcbca05a155d295992ddb3722b1db"}
{"id": "train_072462", "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\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c09e34ee3e56c21c4a66061200743b34e6cf3373"}
{"id": "train_072463", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "9ee1be780ae7ce43a7c19007fb57f969f105ede4"}
{"id": "train_072464", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "33165263f71935f77359e2fd014c12616db926c0"}
{"id": "train_072465", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "3d1bbb2e53f47b4ed6d9bc9b58c4a9ef8c1dfa59"}
{"id": "train_072466", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "12d55912e3b3ae6b857d279e32fad6962b4aefa3"}
{"id": "train_072467", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "10e861a15e03b5051d4f7e16f38c89b43f27081b"}
{"id": "train_072468", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "9d544d219fdb3d761c49e545dd32410b4edce654"}
{"id": "train_072469", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a4acca85eab2641f31cfeaedffcafa20a1f722da"}
{"id": "train_072470", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "4dd0109fe6d2940e1f03d29af2ebb0ac2b04e999"}
{"id": "train_072471", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "355b009b2ef3806523c6f36cd287aeb855a3998a"}
{"id": "train_072472", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "806633ab3696447bd21ff02ef2fc12d7d25bf85b"}
{"id": "train_072473", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f989ffa6f44c844fce6fa28368d2f425855534d5"}
{"id": "train_072474", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "1ece5d139d8a1b2f1fb4cbd6a90a77e3376e80c2"}
{"id": "train_072475", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "04d8cd07d19ac0c3114d8d3125c305546faa8bd5"}
{"id": "train_072476", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "101231df9d3a763bc8aae35115d1e3fc6c3be120"}
{"id": "train_072477", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "c13e125a66350ce85cc7b44b6158b11350f5912e"}
{"id": "train_072478", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "590d1294b42d4bac271a0dadd73505700c842802"}
{"id": "train_072479", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "943e954340a048334c7c5ad96d37c201f8093ee0"}
{"id": "train_072480", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "7c948ebbcb8cad4acac275fe78ae1492bd81ffad"}
{"id": "train_072481", "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\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "da4023c189993160e7907b7733e25d428dacf037"}
{"id": "train_072482", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "cc982c75125277bb1bb31834ad2cbddb5854acc5"}
{"id": "train_072483", "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\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "a14669b41726d6974a366b69874c60fe5e8cbbc0"}
{"id": "train_072484", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "6e0c9bc198eff69c0d7263f3cd848db7d676b224"}
{"id": "train_072485", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "70f0fa3c1a8bc439c653ec2c4d600b6ee0114ab2"}
{"id": "train_072486", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "408051c45e7027d5f7c6deb474ab5f44035a499c"}
{"id": "train_072487", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "602d4eec52ae6ece7396137c415ddc51dbe89de2"}
{"id": "train_072488", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "9925e40919d2978090a4efff987d1e51a4aeabba"}
{"id": "train_072489", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "10efeb6562ac55455023e67150b49013198f5636"}
{"id": "train_072490", "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\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "049690ccfe069c94a1135dd2330558dbfaff75d5"}
{"id": "train_072491", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "cfc7a750c4c753acea3231304368b9ab2662e380"}
{"id": "train_072492", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a6f9446156a731b951e6bb37557bc6fe867799c1"}
{"id": "train_072493", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "e9d59dab40cd8980b36bfe24f1a364f37a499d65"}
{"id": "train_072494", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f7a9453046b3de095a11dc9f6324bfda03084f4a"}
{"id": "train_072495", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "a8d2ab897d690eea009428253e48ab42aaf4ceab"}
{"id": "train_072496", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "173d54ed4e75f02d6383930bdaa2104f9b4316e3"}
{"id": "train_072497", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["performance", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "05b6018ce17a0f036dbae6310a5497170c39fa87"}
{"id": "train_072498", "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\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["agent", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "fb0db58d102eaae5c37069af86f67dfe498e2e71"}
{"id": "train_072499", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "0ddc60fc61864d3962bcdec4e8c88e263fc8adc1"}
{"id": "train_072500", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "28c70a8df8a3c0376fb794ade82d6ff0541db91f"}
{"id": "train_072501", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "497cb74263202b65e76b663b703e1d59c3168be5"}
{"id": "train_072502", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "7bb0aa0ff80c1c00a47f0882081fd37656c4e894"}
{"id": "train_072503", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "49bb9b3672a65dbb27c4f8c93050d04506cb9859"}
{"id": "train_072504", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "7c6480cb8f7ee74c7451ef1cd8e395eeed1fd000"}
{"id": "train_072505", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "853969ca7fa1a28e2b1b53cdf8928db57d174161"}
{"id": "train_072506", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "0e5e5e510f08edd9773424000b8d43ca9b3c58db"}
{"id": "train_072507", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4c50abb83001962463f1692cbfb6df80eed90a1f"}
{"id": "train_072508", "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\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "e3e63f92ca9e542e2c77852330013908974169fb"}
{"id": "train_072509", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "f80c0e7ab238aa7f8f3ef8f372126daae3360d8c"}
{"id": "train_072510", "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\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "d6c65b7aac18b8a0b9257ab9aa7d53def161a8b0"}
{"id": "train_072511", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e8692c5e5f53ca0ae7f8bb2649fd4fdbf08fadd7"}
{"id": "train_072512", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "9b96f8b639c2c2e389df29ef1c33b590fa27ab71"}
{"id": "train_072513", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "7ba9226f459a5eb783337e115c91801cb89eaaf9"}
{"id": "train_072514", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "2f0425f1bd64ab2dbdb3d0364eafec8a59a1f223"}
{"id": "train_072515", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "83cf09b3d7136c7a770650471faf695868e32a27"}
{"id": "train_072516", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "4c3d030ee7abd59074fbd3484a00d445efdf9756"}
{"id": "train_072517", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "af0c820538a184815da87693ce0c184ffe6fc957"}
{"id": "train_072518", "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\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "d8847f03c0800f45b6b7d5eaa315540a131905a8"}
{"id": "train_072519", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e1b78d65a9fee9cbde4a9ece679b373cceb8ef98"}
{"id": "train_072520", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "576dd04c9e43afbe2993edc6d11f4953b5ffadb8"}
{"id": "train_072521", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "969533b1d0d500f9604600e382f3a339ed006bad"}
{"id": "train_072522", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "d3a30eaed64ca69d7a4dc6a137551a63ff508024"}
{"id": "train_072523", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "c58780226ce90b9fbf544d5c84faa1d316153bf6"}
{"id": "train_072524", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "41a064279a7bdf82234a82a9c8962bc3431f5032"}
{"id": "train_072525", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "b84c768b664cee95877423c24225ccb90306432b"}
{"id": "train_072526", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["research", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f658721074960b475f1542647e79fdc03d20b127"}
{"id": "train_072527", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "f2f0fc4497341f1d95826d47b33da2da6c0883d0"}
{"id": "train_072528", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["evaluation", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f3e97f36bf66e896275a14db80c947c5c9af011f"}
{"id": "train_072529", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e2271064d05be32ffa3fc35995798709407b20e6"}
{"id": "train_072530", "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\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "ab856ad8681b735f4c1a55e8aef5cbd5a47afcf7"}
{"id": "train_072531", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "e0407a3abb26065ac1649e4b67efe2702c853e4f"}
{"id": "train_072532", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "75066fa7f3e25e093f5a2f9a50699701b2d995b9"}
{"id": "train_072533", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "ba8016fc4f0a4225cbf0f728ead66327265a940c"}
{"id": "train_072534", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "3d013d439b8496b8afbd2fecbd1115a8526839e6"}
{"id": "train_072535", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "438aac7594eb4d0f3c778d76e3f839fff8e46844"}
{"id": "train_072536", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "9e1d027481754dd080068b56b8221600d616c69e"}
{"id": "train_072537", "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\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "5de52ddf3194776a11dc0c21238af88edda93100"}
{"id": "train_072538", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f16fd8d601d74f19a6cc06b8873d62975b9307bb"}
{"id": "train_072539", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "f715e45cb29c631e73471e27187d4e9a011a43fc"}
{"id": "train_072540", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "55a78e25197544b6df4e2beed67c5d146ed3f906"}
{"id": "train_072541", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "9097afc3633a1627747a90b2688c19c1acfa3722"}
{"id": "train_072542", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "6419db58ed6329fb88c0f561bf5cf1a408f1f98b"}
{"id": "train_072543", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "77c06836e6bed26bf6b3fc3e881b5f2cfccedb01"}
{"id": "train_072544", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a97fe8533b0083f03a7b462eea889bce442efd39"}
{"id": "train_072545", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "30ef96b77b8079c5ddf6f3055621c0cb513c70ec"}
{"id": "train_072546", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "64264f4d3dddb49a4ab71d62305a831325df3031"}
{"id": "train_072547", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "7767ce1fd0fe6356e8752d1467484811e63972bd"}
{"id": "train_072548", "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\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "fddb15027c34b4d0de9300d89c981d37e7d4872c"}
{"id": "train_072549", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a26d367c085e0c10a7cfe4dcf015282567b1ee21"}
{"id": "train_072550", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "568aabe1a3fddb4071d55cf065c4f8f9cd8bb347"}
{"id": "train_072551", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "8be4de57c505b41e69e1e7e06bf7c51725d1668d"}
{"id": "train_072552", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "44bf85e42479cda900745ea73a1c75276a9b4a84"}
{"id": "train_072553", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "ff8ecd35fb875c829d14fef93523b9680c9a4d93"}
{"id": "train_072554", "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\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "1fc0c6c1e2d3f37e762f0d926be085ad928f583a"}
{"id": "train_072555", "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\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "25e17c2d65f9697bc2f135265df9c19614fdcab5"}
{"id": "train_072556", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "ee72d8d43665830810c0cc21515c00bfad855cc1"}
{"id": "train_072557", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b364b4cfbff735f923d7fdab7c2315141728c20a"}
{"id": "train_072558", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "f4163076a589ba938824bc31afd038e35fe4dc71"}
{"id": "train_072559", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "e51f22f6b043346805064f649501ddaf0086fad8"}
{"id": "train_072560", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "b5160c9e7859a5eb4e81470e32b9ea6ae64158c9"}
{"id": "train_072561", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "453668e647333931b5fa77333920b61ad195c802"}
{"id": "train_072562", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "df116e20f019c82807e524f76687cb92d1062b17"}
{"id": "train_072563", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "38193ef683e303d6346f748a6dc2adfd67de8bf4"}
{"id": "train_072564", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "15ac82c638f40d9dd99eda60fde70ffe6e87b4e0"}
{"id": "train_072565", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "c6cf9a3f3bea5b0532a6dcd1f9f69cd67af92c2b"}
{"id": "train_072566", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "0f23cf7c158f6684a45a839852650346b5777577"}
{"id": "train_072567", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "ccc1be1ad3905d5422993d0cc3440a18999eeb27"}
{"id": "train_072568", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "51a2a073a18a3685dc1927d12523140108282cdc"}
{"id": "train_072569", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "dd9adbca35982ab51ebc170a94306f42b3ee576a"}
{"id": "train_072570", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "831c1be2d5c805f1995277d4729feb8733ac464c"}
{"id": "train_072571", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "6e2e48f5a1c640b6881f2dc0797ae5f1d7ef0a11"}
{"id": "train_072572", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "1db480612a4866b831a912796425cc5d75036f80"}
{"id": "train_072573", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "abfa266fa0acfc5049889adbcbc82030aa15fdbc"}
{"id": "train_072574", "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\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "a1af557f126cbd9f1f112183d30459699ab85479"}
{"id": "train_072575", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "57a3708c78fe142984473f72575cbd8b698f141c"}
{"id": "train_072576", "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\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c313a2a09e28b9baa1fbc440c8e87f249a214e9b"}
{"id": "train_072577", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "7d3626e4c605a18c993960fd2af4f6a427b534ab"}
{"id": "train_072578", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "c6027552d83858dd21b5e01500ef4ab7b5ce5635"}
{"id": "train_072579", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "2a1c7867995931ef418d35e364b4c3e763785a2e"}
{"id": "train_072580", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "7236449ebe53b096eae2dd5a916818edb5ee05e0"}
{"id": "train_072581", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "b43abfd5f10d81633fc39cf313e55d6d577e8e98"}
{"id": "train_072582", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "144fce9834737c201451823caa0aa28cccfdc818"}
{"id": "train_072583", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "d47ef5ee6efdae5b2775f9d7c75de3060251ddf4"}
{"id": "train_072584", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "96755d0cc1875c963c583cbd88ae6102994c9e8a"}
{"id": "train_072585", "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\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "d1f624bfa1b3ab33eb0389d91bae0b5b44d30ac2"}
{"id": "train_072586", "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\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "c7188caa6b8c5ad5df423e0ed41f291393afeaed"}
{"id": "train_072587", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "33ad6d890dfb83a06aa3bd76ab19338150541db0"}
{"id": "train_072588", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "9d61ced89efdfd58111f4ce01aebb0b944bfc12f"}
{"id": "train_072589", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "f2ec51c70a7cd8922328a4d84e43e1b2b259267d"}
{"id": "train_072590", "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\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "ada64629b4d38780bd1088baeeafcd48efaeab27"}
{"id": "train_072591", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "70e99adf582b30a8bbe0f36e7291270add81ab7e"}
{"id": "train_072592", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "58637139f076ccf715371fbe9982b7e9bfbb6892"}
{"id": "train_072593", "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\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "c98d0ebcef026083f66c93af9c017c27be40fc9e"}
{"id": "train_072594", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "86b093a2227bb72b707369eb98ae4f71d307228b"}
{"id": "train_072595", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "af3978b6506f351db3d3bfd38d3c0ef4cc2c410c"}
{"id": "train_072596", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "2631ecc998392e08f989f5d890a206ebda750e13"}
{"id": "train_072597", "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\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "c5843a9331afcf7deb2637957c6c079e2c8f1d58"}
{"id": "train_072598", "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\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4f5c98cd2aabd5e43fa0a07d5174c6c62d19e0cf"}
{"id": "train_072599", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "a90e7aa5dd3a922550ccbd8d81a8bb19ad8e3024"}
{"id": "train_072600", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "4af3b5e3d91a154877c7b41c25b304fafe5a7121"}
{"id": "train_072601", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "913a576db1c3f3a3fc3e19259d2cbf7e3b0105ff"}
{"id": "train_072602", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "321b2e5cdd2cfd47828bd70c1684f342871b049f"}
{"id": "train_072603", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "46a360db2d69aab2ce6c1203ecd785a47b7016c1"}
{"id": "train_072604", "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\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "d884c7307c8c9259d71ddf95f2d34b6fde011182"}
{"id": "train_072605", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "d397776fd1d08242c384085570c420eb3ab13b66"}
{"id": "train_072606", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "7e7e009c61fa7284abd9e71bf5799deb9c81ff07"}
{"id": "train_072607", "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\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "7785ea2e8277b140ddc68bb39de684c2bfed3a99"}
{"id": "train_072608", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "4ee3a1e0e5512b9105efba72e613f5597d2d1f1d"}
{"id": "train_072609", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "7dcf41a2b4c910a8246239e2f8e2c518ab068d3c"}
{"id": "train_072610", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "6a0aec13635bcd75069d97640abaf66d564fd226"}
{"id": "train_072611", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "fb1aa25488865ca3895c956ba856d98aa65394ca"}
{"id": "train_072612", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "ec2274d7e3c2d35901e990c64152ecf6369b8208"}
{"id": "train_072613", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "a9756f6f9a48cc18206a2aee27f6f998ffd3b35f"}
{"id": "train_072614", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "172ffe1ee73127e1a7bf9257db0f478e94e2fe02"}
{"id": "train_072615", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["agent", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "ecb227d4ee4c85c7e1308ce6aa2693ad097ed129"}
{"id": "train_072616", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "039f708bb5bee07fe5a6f4774d6d6ee1052505ba"}
{"id": "train_072617", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "81d2346d793e7c83f53442c221399083159eea17"}
{"id": "train_072618", "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\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f65c44f0d7a8cd8fb5348a2ac6e6e0d3fe01208c"}
{"id": "train_072619", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "e6c0764be8d06ea64d923b762a240ce8a6648fec"}
{"id": "train_072620", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "03f83a6aba0fc6f0d805ac31819f9de0f47f31e5"}
{"id": "train_072621", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "630b5a9fb7ce180b7db8b3ec82d0d3a8642dd5bf"}
{"id": "train_072622", "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\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "fa65ad97ac33dbec381b51ca08c198305127e8d4"}
{"id": "train_072623", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4d725e4ea8440e3de73f9e8a3550b7802c68bc35"}
{"id": "train_072624", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "61faf614c8564411fa0d117428efbfcce2f37b1f"}
{"id": "train_072625", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "2dc9388e3b9ecab34bc12f0758ab172bef77687b"}
{"id": "train_072626", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "138508d159569c103261bee4c1b6c64cec39aae9"}
{"id": "train_072627", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "1d42471337c867896276caea718b82c0f9d92db4"}
{"id": "train_072628", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "279daa2aa883a03e263b6369cd875c56c631726c"}
{"id": "train_072629", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["coding", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "89b869c1f5673fc5aaa205560de7bf06735afded"}
{"id": "train_072630", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "fa60158e5fd2dd17741b311d2f08db858bf43fed"}
{"id": "train_072631", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "7fbb10732287d0442db0fd7fbb578f230229c693"}
{"id": "train_072632", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "1ddf7dc2543ab5763679e0ec35131c4c4ec6ba09"}
{"id": "train_072633", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "6b8b0202583ac9f06516844d17301dabab58da72"}
{"id": "train_072634", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "2c59bdf91f4852a1980c59de0976722821c1e1fa"}
{"id": "train_072635", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "67b7596874dd8e2a15d6281fb4c43f3f604b60df"}
{"id": "train_072636", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "5ef33ca35f504af79f74c88540487a62771e8032"}
{"id": "train_072637", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "c6524afa1c68875d5ee91d143bb2cc119d636694"}
{"id": "train_072638", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a4eafd372d8c35db5f7ca9ffa2ae3de8d7e5cb40"}
{"id": "train_072639", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "ecbc9e0192f8d3151bc9042d6a0082f71e3ba0ee"}
{"id": "train_072640", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "7b214346998baf0b7373de99197731dc0ad371f0"}
{"id": "train_072641", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "cacea8cb7f073d082c0c6e4b9c12cec445cdd202"}
{"id": "train_072642", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "41515788fa36682599525fb4a59d14c070d2a893"}
{"id": "train_072643", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "249770db57435a2b3b157a9e7fcc8f6c7e517e78"}
{"id": "train_072644", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "d0c0868b77e0d6e9bfdd7dd0b337b247ee4f7d17"}
{"id": "train_072645", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "0ce2d3c62e32f67774c7c812c8190562655c575f"}
{"id": "train_072646", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a70aa8a69e0cb1acc055f9c5da32748bb81ae23f"}
{"id": "train_072647", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "437d25167ad59d34a263a54311690b38265e9ef4"}
{"id": "train_072648", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "b2f5d30ddc27e9c5710bad6f33ff564ed2374ccb"}
{"id": "train_072649", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "2c251d8ce9a70176ec1c60d6d002607261490573"}
{"id": "train_072650", "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\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "ca94b89cf24ff48fc7cdfe45e95d3f8468150914"}
{"id": "train_072651", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "7ee534ca9c35ff8e731a4f1276e68d728d8d82bf"}
{"id": "train_072652", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "04d8cd07d19ac0c3114d8d3125c305546faa8bd5"}
{"id": "train_072653", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "85fc53b09d7af8c09567edc87b40621a9f087722"}
{"id": "train_072654", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "bd45506e0195f21bcc4b46fbc4a813b085d9c7b3"}
{"id": "train_072655", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "ae5cd47533b793bfab2b36013fc5140469e6eba1"}
{"id": "train_072656", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "8232ad87e1703aa98a0d0a715bf926a0c2704841"}
{"id": "train_072657", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "77fe5c5b9133abf5cbfb1ad78b01b4fc30d08d09"}
{"id": "train_072658", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a4eb859ab7f853e1b2a7afdb6f78718c9ba2af10"}
{"id": "train_072659", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "2b1b1ef1c8aa61e1e76fecce860ae29434427975"}
{"id": "train_072660", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "4244d3768308581a58a07600c5fb11424fd8e049"}
{"id": "train_072661", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "15e6e179d6be13cbf1dd2c8dba7f0bd0fbc3807f"}
{"id": "train_072662", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "75a32e7f3b8db0e43b81e22a8a2caa9cd4b99dd3"}
{"id": "train_072663", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a07028fa4376a0d84412dc8a47b987cd001aff3a"}
{"id": "train_072664", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "7c53a530d33c8c109ba894a1c01e35dede1db6bd"}
{"id": "train_072665", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "173e439712a4eaafe7fc553c83fde12224219a39"}
{"id": "train_072666", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "224f43f3dd065b285f1980b51e2fe9949e51eaf5"}
{"id": "train_072667", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "9a194f006ffbc3a2e7312e16a7ee7e7531903124"}
{"id": "train_072668", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "365124f94d13dcfbc13f7cb2f4385eb68a09bdf3"}
{"id": "train_072669", "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\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "5f14db28b948f90f67f06e80682c1bbd45713077"}
{"id": "train_072670", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "0369319836a6ce66d48c21129f7f4012fd78464a"}
{"id": "train_072671", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["security", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "f12d009fd94e67e8eb26e6d0fb9b3b33a2cdb41b"}
{"id": "train_072672", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "159d9c89bb08bf18de9038654e3df166b6e7c05a"}
{"id": "train_072673", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "b0057c44454895dc5af0a6b785b6f06d46283411"}
{"id": "train_072674", "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\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "a35155f98910b77d34204f4ba9d99666fa8d6ca4"}
{"id": "train_072675", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a42b138983a384ea4ef9327b4b687732eb2b9440"}
{"id": "train_072676", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "172ffe1ee73127e1a7bf9257db0f478e94e2fe02"}
{"id": "train_072677", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "1398118532339cda2938099d9d28439dd410f8a8"}
{"id": "train_072678", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "fe7927178d428fb3ae33dccbab48b1b6fab582ba"}
{"id": "train_072679", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "3c11277c64d5e34033d5cee4e9424da19158cbe5"}
{"id": "train_072680", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "1ded3143dab881d3f7df51867f6583c787248c01"}
{"id": "train_072681", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "947f93f764ad2975ab5b9307345d2951e07c5484"}
{"id": "train_072682", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "dae3c9b7df9793876c0f489ba65481c1893784f6"}
{"id": "train_072683", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "023b5eea301dbcb2b195fdf4d1ce58759bf612e6"}
{"id": "train_072684", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "7b4f9af6cc1fa003fb9d2a6e2a2b7de8430a3ee3"}
{"id": "train_072685", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "ef9450f1fcb77010e96aba3daa09c598f105b709"}
{"id": "train_072686", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "5fca3b36be0d912abcb080e85965d6dcdb93b6ff"}
{"id": "train_072687", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "d755225b68b6f34c7041840248500acf7138b939"}
{"id": "train_072688", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "8e429f9584c33a8211a2d7a100dc352043fe12dd"}
{"id": "train_072689", "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\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "93ceea3a0493136fb4060186775df03f31231d1d"}
{"id": "train_072690", "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\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "023e27c45b3c43bacfb0f40def91b8c1ad683b94"}
{"id": "train_072691", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "898167b565a150cbd559e48052bce7e5f8b1647b"}
{"id": "train_072692", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "9fe9641397be92ef9f562b2b9a2a454b48d66c67"}
{"id": "train_072693", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "7058b3ec0534fe0e0491bdddbdc1858e46cb25e0"}
{"id": "train_072694", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "4e97dee31bc7030d3ce0685791f97e23f8572afc"}
{"id": "train_072695", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "065380916fc0e1e0980fb53da7a3a12797519da7"}
{"id": "train_072696", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "727c47fc57b61afc028576e84504a13a4c499bf3"}
{"id": "train_072697", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["coding", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "7e4efcb5ef97a098e702a1f21ce5dfffa06afc6f"}
{"id": "train_072698", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "31ba9580748468f4a7c9b494a2860019c5f61b17"}
{"id": "train_072699", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "9165749821f61069b109518a4e2a5368cc67a612"}
{"id": "train_072700", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "45c3505a601c417305a8222c482829d4289e1273"}
{"id": "train_072701", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a8cfc8949780908c16cdf75df10eb62176959abb"}
{"id": "train_072702", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "8b27e411e600ad770b85fac9c79f80667b78f16f"}
{"id": "train_072703", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "a69d6bcc70863952109db6af71c89780821a610b"}
{"id": "train_072704", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "dda8c3b61d0a3f29fa314415f50ae063f3a235d3"}
{"id": "train_072705", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "73cf87691bb7c395199cd027ae6e7e6547c8c9f0"}
{"id": "train_072706", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "5dcf3ec20818366bb7764283983c086e00cac6f5"}
{"id": "train_072707", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "d3255abfc306e2a778fa83af2fdd2dd5bd4ee7d5"}
{"id": "train_072708", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "0821b4565a8050b5a6efe5a856ee08196c43d7f9"}
{"id": "train_072709", "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\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "8c44c3c65b753fb0eca24f7d6fbae0d8b66cd54d"}
{"id": "train_072710", "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\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "3b64cf98d90bacefca48ae9c7b4cd9ee1cbdcf21"}
{"id": "train_072711", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "52e5a21b2f3fa2ce5190dc1dbce724ec04798f32"}
{"id": "train_072712", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "7dc51a48bd87ae4dcedd275851bfa2065e753df2"}
{"id": "train_072713", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "9dd6e4246a82398840e1f794d3312d20d9302b9e"}
{"id": "train_072714", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "cafebeaa6674f42d97e3d2e78f03cb20ecbf3068"}
{"id": "train_072715", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "adc954d11411e97725fdf91972ad670dd5996b54"}
{"id": "train_072716", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["coding", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "d1ed87dc1fe5687f378932a827b4aca26386c152"}
{"id": "train_072717", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "94dff5d847d74778a861961ecbf3ca3288664518"}
{"id": "train_072718", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "66633780f9f7bb8792be5771353a12604288c2c6"}
{"id": "train_072719", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "260076bd3c915a6439e17040820dc5a3445a530e"}
{"id": "train_072720", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b54f56556bb7eb282e4a872f6e2a9bb07a10601e"}
{"id": "train_072721", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "5079eaecf65cf14ea534b08ed07eda5738ea30c0"}
{"id": "train_072722", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["coding", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "c22ab1ccbc4699b8905287c3fcfd4d2776db7ce1"}
{"id": "train_072723", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a68ad6e8c8af3d8625dd7b8127793c5aa7a019bc"}
{"id": "train_072724", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "bdc495332ef7637bec2de9c382a7c9261f376c30"}
{"id": "train_072725", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "014ec0132ac411a5c6b48137f6e084c8487835b9"}
{"id": "train_072726", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "cbfe290204d2a70997084b5ebf0c3b561393192d"}
{"id": "train_072727", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "1df053cdecc6f17171297ea92e94f38d9b05c774"}
{"id": "train_072728", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "a47673ca5186e1067c999b7d49df71fc2dda1dc8"}
{"id": "train_072729", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "693dcacd15eea421e81010241a32d4f0c86c14a4"}
{"id": "train_072730", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "2709160c90c8e21ca59701de193fcfbcb37ffb11"}
{"id": "train_072731", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "410e5b62c71a500c301db04bd4dbb018a59efd93"}
{"id": "train_072732", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "bcd3a470f2753ef468070062d87c4330043d8df9"}
{"id": "train_072733", "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\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "eed824d368560de5ac142f0821c269eee221578f"}
{"id": "train_072734", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "154ae6b5f256e0eab7abb81d8cb82f56392be224"}
{"id": "train_072735", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "b862a66fd5f4719c1702f82b183afa221e7f60c8"}
{"id": "train_072736", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "c13e125a66350ce85cc7b44b6158b11350f5912e"}
{"id": "train_072737", "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\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "987a7fad340ce22d3d5ab5b4037116f4da54937e"}
{"id": "train_072738", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "8664758a35521975ad7de481bc74024dd5aa166c"}
{"id": "train_072739", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "26410ecbfa4cbcb13396d0a17e478af0de859fa1"}
{"id": "train_072740", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "59440bafb6c6624388e696653a4b6e696d46cac1"}
{"id": "train_072741", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "c716715859fa4f18eac6dc579f6faf4a6baf735a"}
{"id": "train_072742", "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\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "3f41ce15e096bb712b8ca9e0fbc66ff619b5e27f"}
{"id": "train_072743", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "da287370cd4be8a69b94240c39c90493f4d986d7"}
{"id": "train_072744", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "ff483c90c3528ec331ad55aa05d377b06546edd4"}
{"id": "train_072745", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "e5828a6c9c9a8cc73d83de90220af74c8446c661"}
{"id": "train_072746", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "59750239f0b9fcbaf29965a609377d59bf472f36"}
{"id": "train_072747", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "33d90a954a4b8bb63f8db0b0809ae494d220695d"}
{"id": "train_072748", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "4d867300cdcfb9cf74939e78ca7cee3b333eb21d"}
{"id": "train_072749", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["research", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "73e6657c630cf2ef65e46d6950ea096ae211bb79"}
{"id": "train_072750", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "5e1c833b711b93643a93ea6d44ed548a1804ff01"}
{"id": "train_072751", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "c067bdeed9d06141c55f87fb4daf861d8290f3d2"}
{"id": "train_072752", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "7ec31525665d03cb77061041cc4e9cd28c373155"}
{"id": "train_072753", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "3bada638ba3e4bcb5bcc00684e748573ccddf1fa"}
{"id": "train_072754", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "98f0c001559abb28167e8f55d917d0d8efb36bf4"}
{"id": "train_072755", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "404c55f8b985c5d8dbb19d82aa9555500ee564f2"}
{"id": "train_072756", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "286188b8549f9995b6781b61e77b5e0e3ed79140"}
{"id": "train_072757", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["governance", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "2ec29f8a21e1edbdf549cab42e1b87f033f4345c"}
{"id": "train_072758", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "a428e4f33c1b3a67731ada39b4a61e654a31e4b5"}
{"id": "train_072759", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "5aa110d3a8564e1a6e014748df0c7e8a8574ca09"}
{"id": "train_072760", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "3fd46854afd8159c519159e73dcf1a8c0a313db0"}
{"id": "train_072761", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "8bdbd802fa4d24c7c9fe84291d5c1fa1eeaa3d65"}
{"id": "train_072762", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "1a95a8f5745c0e4de4a4ce97e8f38e09eaa7dfd3"}
{"id": "train_072763", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "ac55a1cc5ec8bc0d834b74017dd77e38423d029c"}
{"id": "train_072764", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "ce22b8b51f5c53966839c91842be5efac1bf16d6"}
{"id": "train_072765", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "8d3855e332e9b440d3a161e80284ae3665468f3b"}
{"id": "train_072766", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "4012fb4a145746422fa4a20e1974b6999e444a45"}
{"id": "train_072767", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "cf4b3c00e341c098ed41f480fd9dc64c7a53cc48"}
{"id": "train_072768", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "b5432b6069bc1908a552cb6fbd2912da5517e5f8"}
{"id": "train_072769", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "e31a721348c312c04ac2b9c77a62a8cac1ec38d0"}
{"id": "train_072770", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "c6f25ac90171d8731c10239a1031efcac6f87ae8"}
{"id": "train_072771", "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\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "928936c073ea59dff16afa0d1f3a092ac96d89ee"}
{"id": "train_072772", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "fa504d982b7d6f52785266e48348681e8c4988be"}
{"id": "train_072773", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "12a2cae999e41ad0f183afaef8279bfb98295126"}
{"id": "train_072774", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "c398013de3dfb709c6f51531f37a9e9a84f1a2e2"}
{"id": "train_072775", "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\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "fe56adc6b1bcfe520a10d99e85f9e373bff3362d"}
{"id": "train_072776", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "6de8c11e98dea42d486a3c7e100197b50c74b681"}
{"id": "train_072777", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "e4bdf19152541a61346c1354bc6c157d8d98f579"}
{"id": "train_072778", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "f4861990ca274c58f0904bb0706d81402225c778"}
{"id": "train_072779", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "619e0efa41e487e5fe18e0f013dd9feedf7772ff"}
{"id": "train_072780", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "49e2b8da27d5b1efe9d642735ac897b45b54e6b5"}
{"id": "train_072781", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "848bde2a4b5ba786b14d5e58fdadf3d5d585f9ca"}
{"id": "train_072782", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "a09f5498f5ecdbe10d64ba7400eeb3ef2ca3bbd4"}
{"id": "train_072783", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "a81f3a329a8b82283325a847dc727d3c2a25e7e6"}
{"id": "train_072784", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "86cc2f726fae292e3ab00687a57608c649a345be"}
{"id": "train_072785", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "93e7a8bed7521f8433028b7d13530524fd03f0aa"}
{"id": "train_072786", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "099083244e40cc204a92642a709b826d79e4a19b"}
{"id": "train_072787", "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\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "2c17c5fbd63cd835c284d44427e26bef8178352f"}
{"id": "train_072788", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f74a406c59a1bc28bd9f2a2dc7d9e35e60c43e0f"}
{"id": "train_072789", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "159d9c89bb08bf18de9038654e3df166b6e7c05a"}
{"id": "train_072790", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "7d4c37ba9b79c30c636478373f002191048e6c2a"}
{"id": "train_072791", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "3b1e61fc5b927813a723ab2ebfc8b81d865fc84f"}
{"id": "train_072792", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "4c03a868799e1b5e7effdb2147b01200a470296c"}
{"id": "train_072793", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "286036a8ebc3de6f1d407236a46d87b579c604d0"}
{"id": "train_072794", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "be24799ffb8f78c4bb0baabb4c9b1517d72989dd"}
{"id": "train_072795", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "a7083f2bc7a8c239e3fedbdf1903166bd9b16038"}
{"id": "train_072796", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "914e486a68cf61eff1ffcbcc5c6369c7ce771a75"}
{"id": "train_072797", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "a0a214394b68bc9264ef0c2232177ee8409e1924"}
{"id": "train_072798", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "daabcb96fba793772c3387caa63a96f9d69ee71a"}
{"id": "train_072799", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f939dde43f35c6e4a0be1b8b13c07f5d3ee97201"}
{"id": "train_072800", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "b17ebbeb1e8f571cdc634d3e951a2329b5e3f0cb"}
{"id": "train_072801", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "604e392eea495aa3feeb3aed2436fb588109380f"}
{"id": "train_072802", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "89c4966d30201abbead2807fe50695be427ba9be"}
{"id": "train_072803", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "d27bac1d610f601b046157664d75b30ec4259860"}
{"id": "train_072804", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "4a48f95524ade2a2e0ead54d4b5eea0d9d2c65ee"}
{"id": "train_072805", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "16f4181870f6c40f9570e557e2d8c7ca2eaa9c4c"}
{"id": "train_072806", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "df044e4380e0b30d075d57cd90f12aa57ffb02d1"}
{"id": "train_072807", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "bcae1d93e5451e896afe72547876f8f006f186eb"}
{"id": "train_072808", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "3ce1c8ac2a9b827cca3bd9d50f99f1b2f6bd5978"}
{"id": "train_072809", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "1996669c3cc19948f62d50df269a7de3e76aa881"}
{"id": "train_072810", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "b8b910aca9a7081194588b08c81869f4b092cf02"}
{"id": "train_072811", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "0e609322e07e53098effa126d38934300b0a06e8"}
{"id": "train_072812", "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\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "3feec814de6508fdb7fcbecf032a64e923cc96ea"}
{"id": "train_072813", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "576900abe391849ac32f26997c9eb0b7772d28ba"}
{"id": "train_072814", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "ddb5a4aba9740c8831b435d7313cf7904ee6ac9d"}
{"id": "train_072815", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "154a1697ef77c583c8004465ffb8c49ab273ce50"}
{"id": "train_072816", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "093f0a478260c6565167948f26ae503912764c7c"}
{"id": "train_072817", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "d511452825bb7959e59f9e1ac76540a1a2b2c062"}
{"id": "train_072818", "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\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "3feec814de6508fdb7fcbecf032a64e923cc96ea"}
{"id": "train_072819", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "eee1145211863ef251fdef9cdbf898bd3906b2e6"}
{"id": "train_072820", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "06e3719080f9908804ebc01594a316a637707011"}
{"id": "train_072821", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "6243243945e8b1e2692bfb6c268b36c2e72c79b8"}
{"id": "train_072822", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "b145872c5e587d4f084b751a9a100fc85b013767"}
{"id": "train_072823", "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\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "3ab7ffd25d01ee6d810e4e9d01c5800efba90e77"}
{"id": "train_072824", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "fc73a23fa58ba71e6749840dbb92e13c7aad4907"}
{"id": "train_072825", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "2fc505308aec4467dc0461b8ef6a266b7f59fc2c"}
{"id": "train_072826", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "39707bc5e84a98352b84657ff08e85560790141d"}
{"id": "train_072827", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "d25515a77dc6791bed56993c644965c8b70c56fe"}
{"id": "train_072828", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "dda724f4f3acb37a8e025b35841baad288a9b25a"}
{"id": "train_072829", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "7db8f396864232e6bdbc19c7689e0c2e92d09b97"}
{"id": "train_072830", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "32733306497a9a8160c6b8930357e3d8be9450d0"}
{"id": "train_072831", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "6dcc9ca796c3a8b7d7179798b36ac7ed46dcbc0a"}
{"id": "train_072832", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "523bf72a8fbe0c5e0d11a3d31a4915c82e93cf21"}
{"id": "train_072833", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "7299effc1c51079cd0baec41d959932883211c84"}
{"id": "train_072834", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "9d83a1bcf93de0238c205683fbff0af488f377bb"}
{"id": "train_072835", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "5d7f401347269fdf00b75f9b137e0019974582e5"}
{"id": "train_072836", "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\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["performance", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "78345ccdbdd8dddd97cd4eb31728c4d7197e47ee"}
{"id": "train_072837", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "92304c8619e170b4061e5fac0309e6251b3aac0f"}
{"id": "train_072838", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "5705ef3eb4ad3a5941460535e813e238c10fa9a5"}
{"id": "train_072839", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "4681bb93f966b7c7624c84bdb08b662cbb0564bf"}
{"id": "train_072840", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a54a43d56de2fe9bfff0f554d27c776444bfec6f"}
{"id": "train_072841", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "6a88082401a612ba4382dd372dd66b825ef11abe"}
{"id": "train_072842", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "9e3fa9a82b953d6ac33ddad957e965b65871c8df"}
{"id": "train_072843", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "f67313aef7d53b24ab20cdc15b38586ba0659ce1"}
{"id": "train_072844", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "a992bb19ea527b99b3f51fac304583cb5def0ad8"}
{"id": "train_072845", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "0ec0e424d2b2728e508e76c9c5f12720c40cd150"}
{"id": "train_072846", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "109a04a727800c2635da8582f97e2c13128a5e5e"}
{"id": "train_072847", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "8dcaec8af20436e2bde32490b89197f2c80added"}
{"id": "train_072848", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "f97ea717caa30943a7bded196b55b56f9336c82c"}
{"id": "train_072849", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "71697016981aaca265eb6792d95f9948f4d64ef3"}
{"id": "train_072850", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "cc2c803bad070c72a56b31a52c863ff7bd1a7f30"}
{"id": "train_072851", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "c5224e97a542c708ceebc05bd3e96e9561153a7a"}
{"id": "train_072852", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e9d3eeeb9549715841ff508c31adc4528d2005df"}
{"id": "train_072853", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "d4d4b2070aa2c81c2c53bf0620d63bb209cfc5a4"}
{"id": "train_072854", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "a566fb8a39d1f1624818fd322ba419dfdb70689b"}
{"id": "train_072855", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "69d7e19e747a860c585703f6be0509a67326a667"}
{"id": "train_072856", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "45868e7a50ca0ef0242888196af73c6c2b603c44"}
{"id": "train_072857", "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\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["performance", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "3f41ce15e096bb712b8ca9e0fbc66ff619b5e27f"}
{"id": "train_072858", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "e9f7a1f5267385da4c4a8daa3aa6678ddd5ecca3"}
{"id": "train_072859", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "771bebd665244f68f8f7c14ff4278bf8c423a8d3"}
{"id": "train_072860", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "b527bd368aa106a8320746f63f06f4427a4e1ec3"}
{"id": "train_072861", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "9248d23334fec512fa589f45b5c2539c878dcebb"}
{"id": "train_072862", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "1b4da5c481ac1fe414b0671561937a0ffdb3955d"}
{"id": "train_072863", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "438aac7594eb4d0f3c778d76e3f839fff8e46844"}
{"id": "train_072864", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "e777da60e3b1e6491d0494835b3deb605b266b12"}
{"id": "train_072865", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "51353ca582ac261b78f3208a5161be1b70a3fd5b"}
{"id": "train_072866", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "d84c11e6e4217f29ef7cf5c13115b7378d6be221"}
{"id": "train_072867", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "8eb7697a47afd91657cb1b85a8730ee07c32cd10"}
{"id": "train_072868", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "417ddb70a9166954ecfe95c87a8df123586cdb23"}
{"id": "train_072869", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "79299b843b91f41f740af0d3e410648c91050a31"}
{"id": "train_072870", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "cc5fe6df6d67fabc7c65e482a8ec8c1b8c5a44bf"}
{"id": "train_072871", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["performance", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "3a891138cd4dae86b9d1004e195a993f47048459"}
{"id": "train_072872", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "ff3d87ef688ea2e616ff1300bd6e332f55e7faca"}
{"id": "train_072873", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "c3ddd0276fbddfb7f277871afbfe6bd65eecd2c4"}
{"id": "train_072874", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "52c5f2273c5c6b3c29857afce1e964325f6ec6b2"}
{"id": "train_072875", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "5376870639793ef26e3da13b67e5cae209c4cd4b"}
{"id": "train_072876", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "94b0099aec5d3f70006f850919a52d6018724303"}
{"id": "train_072877", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "2d8a0bfea3a3d06b7ddd3267510c62d313862ba0"}
{"id": "train_072878", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "ea79651e8dcf292f82b80b7a0dfa6295859a8de7"}
{"id": "train_072879", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "b1d058e4522d9a3b5eeb051ad88df235e6f1321e"}
{"id": "train_072880", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "a64bc72b99af78cce7b855aa5cfc7dd40ab1c00b"}
{"id": "train_072881", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "dc8e01ae5d3d4d805a20cf9ca04c6c4fe513614a"}
{"id": "train_072882", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "05989c31f40a7601c220fc55940ba073b684ec9c"}
{"id": "train_072883", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "6073ee9db4be70b2b39bc05b17b22dafff13938b"}
{"id": "train_072884", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "cb72b2c5d24228e720ce6f1e75d1c19df67fd683"}
{"id": "train_072885", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "38c81a1af994daca76648cc96bbe12915c80c920"}
{"id": "train_072886", "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\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "99263976f32c08074d8b04b3b11bb0db6c9591b7"}
{"id": "train_072887", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "dce167984f4d7f54b19837e0d43bfb9504e9f88d"}
{"id": "train_072888", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "dfa2f35a565f442114a794caeebe40f33b1bc8a8"}
{"id": "train_072889", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["security", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "df131fda9dbc7c2ecb1dd9ea16fbf8ae40512e15"}
{"id": "train_072890", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["agent", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "00ed7ac449119ce690ca732e044329abc37249dc"}
{"id": "train_072891", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "895d8f33cb6f06508e6f7c1928200dcbe8cece5d"}
{"id": "train_072892", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "55ac8e868de881c5f4e444008b58ca2a2c053aa8"}
{"id": "train_072893", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "688b6529241d3568c97e57e452a1a4e2111491e7"}
{"id": "train_072894", "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\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "944c774a3b5829290d363be30b61e26ee5e9f3ed"}
{"id": "train_072895", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["security", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "eb06134a291bca1fb8ae74bd64bb659ff55a0a36"}
{"id": "train_072896", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "14470e25565b0c22822ae35c7101b39ef2ef0ad1"}
{"id": "train_072897", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "21022fb3c1da2f03551dd9b21a70787e4b053593"}
{"id": "train_072898", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "638bef76db5270ba9a07f780a5d3f07194133931"}
{"id": "train_072899", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "121c9eba19bef39355f79a4d9acc4bc792a154a1"}
{"id": "train_072900", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "ece07dc8edb94cad198d6df36d8ab30f4b7f56f1"}
{"id": "train_072901", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "8d4b4bfe6e9d2900c6388857688226402ca851a5"}
{"id": "train_072902", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["coding", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "ee5deb8649614fa86cac03bda12ef2ce829e4dde"}
{"id": "train_072903", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "72aee809eef90ff8a38865963abe9906963e646b"}
{"id": "train_072904", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "0edea10f4c8e3772490accd00d2fdc319454bfbe"}
{"id": "train_072905", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "316a5cb83bdeac456020a641610f1cbf04a1699d"}
{"id": "train_072906", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "91c8d404a5abcd9c8e722f6e99b2759096f013c4"}
{"id": "train_072907", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e1983414015c0341901ed354a03a1b8812a92d19"}
{"id": "train_072908", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "066f682ee748d718aba963751911e6dee0f7314e"}
{"id": "train_072909", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "e6099f69bd4920a54f067f67cbabf7484c347b11"}
{"id": "train_072910", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "dd86698f68648042d505f6bb290b609a60045f3b"}
{"id": "train_072911", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "164e205f8b4cc0ad640be202c249839bfcffe59f"}
{"id": "train_072912", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["research", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "b17db9cc24d9e84fd7f2b3edc1529388df74ecbc"}
{"id": "train_072913", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "b4eafcec1cb9400c7939ac966672140e3c671670"}
{"id": "train_072914", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "5538edea2d380aae9aa9917a50048d02b3e031fb"}
{"id": "train_072915", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "7920f9091e787a78452df30eacb16291244796a2"}
{"id": "train_072916", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "46419214383bb63e9e575f72f2846327f9b3c5a1"}
{"id": "train_072917", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "579b791390b0f8df3225a40b37422b5afb21b91e"}
{"id": "train_072918", "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\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a8304293f0689e52e8ef15efb1754124be992c17"}
{"id": "train_072919", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "5debd765f9de29ccf6e5fb0e84f19930a525fa6a"}
{"id": "train_072920", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "63f5fc8e110c2651d97ad904b58c1d07aa32e574"}
{"id": "train_072921", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "147421f8723ce46fe230158a9501fe8bcf960db4"}
{"id": "train_072922", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "db1ddc4708db4aa52a4ed8dde66deb822f4e82b0"}
{"id": "train_072923", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "857ec5be3c888f4204ebc2bc50e1fec3e73fca8b"}
{"id": "train_072924", "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\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "2f7f3b4e1c7dbce55b8dd12d95bfca5383dcd053"}
{"id": "train_072925", "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\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "b68aa8d6a99e9262d8dd62f20bace6f06a78bb96"}
{"id": "train_072926", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "fc4ba7a317d24b7c0dbd23167f635564b63639e6"}
{"id": "train_072927", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "016ee965bf128ab58386bd98e14d6761f58de5e7"}
{"id": "train_072928", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "4a6a0ef34ca24250df857e5ddfc632826a3c02ce"}
{"id": "train_072929", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "d6e9e2bf38e00706012ef96e789eed45abb5c2e9"}
{"id": "train_072930", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "01c3ce1e07faa4fb9c3d1603ec4edc2f06371b6f"}
{"id": "train_072931", "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\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "eb0db8ffb2c1d3c92784289818544e138f174a3a"}
{"id": "train_072932", "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\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "be6dce8fa9d465c97c4a7bb76255c00868bbcb6d"}
{"id": "train_072933", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "72cb5844a05ac23b91c6e6a216fa473f42c6b029"}
{"id": "train_072934", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "679555ce63be3a66be792c03d93ecd36d5c924e1"}
{"id": "train_072935", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["research", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "a261c5002fdf2bfd97390fcb5fa2e1f16ef7693a"}
{"id": "train_072936", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "fbe02c19deb701725bb8b3060e8dd6353654d2e2"}
{"id": "train_072937", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "c9337bf7294bea9c6cd215c4bd1463148c5bd25e"}
{"id": "train_072938", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "e8db5bbaaa8b1468d11e72fcd122614d8ff047b6"}
{"id": "train_072939", "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\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "302ef3bfe76efc1ad67a3886cab9caa5dd0c3dc6"}
{"id": "train_072940", "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\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "d1f624bfa1b3ab33eb0389d91bae0b5b44d30ac2"}
{"id": "train_072941", "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\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["security", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e462fba52c0d3b3b543ef5351cc21a0e0c552d36"}
{"id": "train_072942", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "83f60bc4e8f3b1ebe94d17f7e0c22a91089673cc"}
{"id": "train_072943", "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\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "7a27415268eff5952b27ec375efe60a94f134ce2"}
{"id": "train_072944", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "1d9c14923cd46c12a1dfeb73ffb26facc3122e60"}
{"id": "train_072945", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "97efb7fcda1a55c0000db9c8e1dae78dbd02d449"}
{"id": "train_072946", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "a05a3280fb14ed280de5e5cca67d6f95dea4d65c"}
{"id": "train_072947", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "731c4d583dab8d534496804b7445dcb6015685ba"}
{"id": "train_072948", "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\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "884bf890f028ac74011d2cbc3623cc7962b11fb1"}
{"id": "train_072949", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "eb965413aff84fbe32410faccdbcf56955cde9e8"}
{"id": "train_072950", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "56174617c7292e0dff3c8ac45386c1b275f3bf3a"}
{"id": "train_072951", "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\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "021b6817e5323056a4be43e92d6a9c14c35529f9"}
{"id": "train_072952", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "31d411b1777ba2d62968b6396d381d9ed9a2218e"}
{"id": "train_072953", "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\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f044cafdb2edcffcbcd89db3a9b680c2ef3dbdec"}
{"id": "train_072954", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "0d2e9005fe85971c145ea07364d46f3906ac748f"}
{"id": "train_072955", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "c2eb1ef0389131ad77684588804ad02ca6ff374d"}
{"id": "train_072956", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "6479f1d1ce203a42d5695cd41bac2170c518cafb"}
{"id": "train_072957", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "3f0fd682929496792bb7768058d9bb95a53b61ef"}
{"id": "train_072958", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "d6cd6059c1b00e346665ea366dd85629cb09904b"}
{"id": "train_072959", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "14dc1b180d4e7eb3cb7af540066b7fbd36efa916"}
{"id": "train_072960", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "5bb553e73cd9412a0813a468f236f23ff4dea91d"}
{"id": "train_072961", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "da34f0258e580105ee9f3bc74824dcf2614b37b6"}
{"id": "train_072962", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["evaluation", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "a99062cc419d9b5a0c44f59914cfe70eded945a6"}
{"id": "train_072963", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "2637586b8386e4a44f52525770148f5cff0b14a5"}
{"id": "train_072964", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "8e0dca0b7ba9dd6362686c6d70382dfb9c3ce2fd"}
{"id": "train_072965", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "ad4861785e8475488fa5ed3e9408a68ed7adc6bb"}
{"id": "train_072966", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "7bba7f52b558044805409671dc661e1cd3812a8b"}
{"id": "train_072967", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["coding", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "e80bc915cbe725e8be788ba63ec54dc868f73e7b"}
{"id": "train_072968", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "db55fc88461b4a5ece608429b8be200d179fdf38"}
{"id": "train_072969", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "354083d148cc9dd93976843f62ef5d65f586acd2"}
{"id": "train_072970", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "562445e7206bee127bd2524724dabde0a1ec022c"}
{"id": "train_072971", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "a62b270e25268ebc9fe750cafc38886931d73195"}
{"id": "train_072972", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "bcef57e227aeeb9decb777c1e54f0b1d84e43854"}
{"id": "train_072973", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "072b41fbcf23aed5786ee3c5b0702ed1f7c988b0"}
{"id": "train_072974", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["coding", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "0f0d60f8d1ff2af9de0531639f05847dfd5e9fbc"}
{"id": "train_072975", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "73264b7523e3dce73451278f7c6a72df082570cc"}
{"id": "train_072976", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "cd1a082ba9209c3062c487b687653932a106577c"}
{"id": "train_072977", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "7820aca5fa808950243a32a510b4c543c60eb201"}
{"id": "train_072978", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "470fc57316b9ff975aac8157f62167ba69ed4a18"}
{"id": "train_072979", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "bc7d7a9e4d8bdb2ee98f76fc0e391097c6db42ed"}
{"id": "train_072980", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "29be7b6a6c59acc4d32faf3214dfcb29a6e2fb3c"}
{"id": "train_072981", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "2b09cd7c5eca8887d723e3f7e46b475bfcca871d"}
{"id": "train_072982", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "cb045b3f32eb71efd938b3db3d001ff6067b2b9c"}
{"id": "train_072983", "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\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "09de2a2c96b5b2aac1b95f016c45908972a4889c"}
{"id": "train_072984", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c62d4f14cbaad633cd35526285e3ed5f947bf3f1"}
{"id": "train_072985", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "8c687060064fcca482987b4e88d1efc5ae07f3e2"}
{"id": "train_072986", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "cbcec364ed384ced79028c2239e7e2a957d020e3"}
{"id": "train_072987", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "cc8a4a0f4e28241da0c5a7742577d9bfdbe9d590"}
{"id": "train_072988", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "8720dd8b0ed3ce5f2fc52f15e3d32ec1ba057140"}
{"id": "train_072989", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "86c13d6169085eb1104bb64e4d79a04684696ddd"}
{"id": "train_072990", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "f42f0c350677cbedba664d6d8cb31d65b43f4eb9"}
{"id": "train_072991", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "deafccee74d1b9fb79e9d6937bbbd4eee9d9f821"}
{"id": "train_072992", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "17974173bd1bd9dc73eafdf47990521403507703"}
{"id": "train_072993", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "9cf62812d6b4945192ea65a7e5d099864b978572"}
{"id": "train_072994", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "4b8cce441765b488888b191e53c4f70222ef3685"}
{"id": "train_072995", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "8fd7c6c1e87d97fe7b78b599fa4bf2a7d85101f8"}
{"id": "train_072996", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "1ee4a1cb399af302acca7deda09700d5e73d39e3"}
{"id": "train_072997", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "724b3f8e4f63c0fc2b81a9f4219e5d51601e7363"}
{"id": "train_072998", "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\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "6a8a8025249692ce45f976ec01ccd35f2a54186e"}
{"id": "train_072999", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "3d5290e640f4cc2eebff7dd6a7069aa9a7174df2"}
{"id": "train_073000", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "8907c41f67ec18dd49292ab809eddf0bb0739887"}
{"id": "train_073001", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "c8ee561fcc9fe7147a84910b1be09941fc118485"}
{"id": "train_073002", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "17292f089361dee6cd1b7a1fc87c869eabe38481"}
{"id": "train_073003", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "2bf5e05cef1aa1c5a3e27df4e9959deb4b301df2"}
{"id": "train_073004", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "6bef381cf87530b692dcdc20da28f63ec966da4d"}
{"id": "train_073005", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "bd1fd3cdbc932d0863bb4c120fa47e315c371237"}
{"id": "train_073006", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "da3dbfa184fca3f8f914df0da0dfbbe7c3e3318c"}
{"id": "train_073007", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "567afd009f5a0f47c534c7e00524a081dde27795"}
{"id": "train_073008", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "fd3e18d344c96e6df6e6f9cd145f4b858fdf804a"}
{"id": "train_073009", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "7ec31525665d03cb77061041cc4e9cd28c373155"}
{"id": "train_073010", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "889b9753d9b25c6ac9c6930fb104713d5715bf2a"}
{"id": "train_073011", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "afec8172a1669851b8742a2a612491da6662bd5b"}
{"id": "train_073012", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "2e2755c2b1958b8bb6fb317b73b4de5017ae3a11"}
{"id": "train_073013", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "c9e8081ccd358ec331de8ef68d4bcaf13e74c695"}
{"id": "train_073014", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "42f0759037448f70fcbe256ce804f904c61c0a69"}
{"id": "train_073015", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c962a028f02a598da79ffa3dcc3178b446d0133a"}
{"id": "train_073016", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "a4765e3c6d21c9e1291fd7f523da0e00d45a2e5b"}
{"id": "train_073017", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "a2b04111458c333ad8c1bd8c3921819a93f36495"}
{"id": "train_073018", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "ae26be2023d8fd4f7013171a764dcf14d0321cdb"}
{"id": "train_073019", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "0c7fc467367c957b711343eb47954707b3c258c6"}
{"id": "train_073020", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["research", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "c5cba086148a4fb2339a34237ae0e091d7cb388b"}
{"id": "train_073021", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "877b7113b0bcffcfa23281b1af5f5b7bef5558a0"}
{"id": "train_073022", "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\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "1aab72284b4fee0e9b821117d790bd365ec0c683"}
{"id": "train_073023", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "bbc971daf3983d139ed50a1c168043ff5828d0c8"}
{"id": "train_073024", "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\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "8bceec01a16578064857d62a7e6ec327b449e7c3"}
{"id": "train_073025", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "dd77de8b5d5cc8bae351375e54f50882c2fdba28"}
{"id": "train_073026", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "64bed467e608ac3700b8af699ac704edb93a6fa7"}
{"id": "train_073027", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "b21f2a539472369c7e938c1244d3ed9c16ba0f98"}
{"id": "train_073028", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "5c515f7cf9211171a1d24df23c981b2b198b2d4b"}
{"id": "train_073029", "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\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "ede3a8bfd50fd9e889090e9a54969afd1ad68d63"}
{"id": "train_073030", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "b27123724a5339c4f9363f6fc59b5747deaf5433"}
{"id": "train_073031", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "06a3366d2197a60939fae40a583a456cb480905d"}
{"id": "train_073032", "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\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a29a68d2bb5c4f20f60b93305f9d262f3c6a03d4"}
{"id": "train_073033", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f48e35c77c72e2cdea9734b8d5aa3384b3eb2429"}
{"id": "train_073034", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "486dc328f406fa576cef1d833a3537b6eee46065"}
{"id": "train_073035", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "75bc660e35f6fcaa46730fb1745dc6065b313ad8"}
{"id": "train_073036", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "55a78e25197544b6df4e2beed67c5d146ed3f906"}
{"id": "train_073037", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "bdf6dba2cf9b0feb041fff864b1ddf415b9f12ab"}
{"id": "train_073038", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "b88e27a3d2e44a7bd7d97e381f438687578dbef7"}
{"id": "train_073039", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "070e188c6570158f8d63dbe847d312397e9fa24e"}
{"id": "train_073040", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "c348ceb630419fe9c141f1a006859750198e2fba"}
{"id": "train_073041", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "ce008b1c55e50ead506b501115492062f89ac815"}
{"id": "train_073042", "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\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "8b652865c229c31a9c22635dc639881216515bed"}
{"id": "train_073043", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "6289e46c9b6f66ce5b0db70d4f4c751b94f48eb2"}
{"id": "train_073044", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["agent", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "8e4ceaeb168aae5ce18481d73af88ca054f98582"}
{"id": "train_073045", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "8aacfbcf4a6d1aa4af7bc143abe05354037a331c"}
{"id": "train_073046", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "cf3f565fb1bb923583eaab5cdd11aafdacf746ff"}
{"id": "train_073047", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "ad60217dc60acfcdf6ea89366fbf93ec99509a65"}
{"id": "train_073048", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "7b17dfe8535df15bc842d9e7a21a68eb61840317"}
{"id": "train_073049", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "25fa0903e3bcf3c1181325fe325fe1e407c138f3"}
{"id": "train_073050", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "bb7f2df13db407c5f7646da98304d7b38ab41def"}
{"id": "train_073051", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "66df9f989e1a5065439e599c7e060b000287c139"}
{"id": "train_073052", "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\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "ec6cf59ae9f612a2763d1296b0418f57d5cca4fa"}
{"id": "train_073053", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "df116e20f019c82807e524f76687cb92d1062b17"}
{"id": "train_073054", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "75d397f4d0c1a786c8cf120b52382f29ed7f205b"}
{"id": "train_073055", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c1fe18ad108caf540cab3f17383d1d7ab09b9094"}
{"id": "train_073056", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "688b6529241d3568c97e57e452a1a4e2111491e7"}
{"id": "train_073057", "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\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "1b3ece31e5ce539d773f944501f16c7e587fe0bb"}
{"id": "train_073058", "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\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f0ca4186e1691b46794185c00fe84a1e7d27c095"}
{"id": "train_073059", "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\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "5aa9ab865ba54b52c68de2b63975146e24700a6d"}
{"id": "train_073060", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "231795ef8eaea72672456197948ce3b1d8b0a4c7"}
{"id": "train_073061", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "40580a70ba24445a3eaca4a04f67c6e1bd6b00f2"}
{"id": "train_073062", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["agent", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "3683c7705e6951a0d973b8afbf02d5ce71cd7b71"}
{"id": "train_073063", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "e6d1a9b0cb6032da75812ef1140e25d06efe990a"}
{"id": "train_073064", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "5872c5fbf902c4e5694218985bfff5d07d5c9af1"}
{"id": "train_073065", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "74bfb5c1b69ecac24cb5c916f03bdfca01dcfc4a"}
{"id": "train_073066", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "9f7551d345d397ecdd06601894a643749dde5516"}
{"id": "train_073067", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b145872c5e587d4f084b751a9a100fc85b013767"}
{"id": "train_073068", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "bcbd3aa62af828dd086d48baaa530fb6933ceea1"}
{"id": "train_073069", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "874fd218fd3d5dd9b95bd30bb3ac6922f2241393"}
{"id": "train_073070", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "3f372768e95c7ad6b286b6c803d234bedc782072"}
{"id": "train_073071", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "bdd5c84d523f32aa3428375d74b51cdf7498eee1"}
{"id": "train_073072", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "6c804547d12284c0df859cdc3508fa022a154de8"}
{"id": "train_073073", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["evaluation", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "d8a6a4c4f5d0b99029602747e6b4fc6a8237040e"}
{"id": "train_073074", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "b6695da613b152d1273ea1c221ff54d3680a7585"}
{"id": "train_073075", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "1c7b131f3faa7702c356d5ba938f37b2e42ea432"}
{"id": "train_073076", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "de60a364d6e6dc7def2621e8d1729b3d846fe430"}
{"id": "train_073077", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "cba3fc2c6b929ee7939357c67bdbb540030945ea"}
{"id": "train_073078", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "51c7b1d25973fe8dab607f4e80a317bf07291d0d"}
{"id": "train_073079", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "33330e1315bd450b9757849902cc365375bcdbdf"}
{"id": "train_073080", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "c16db4dd542d681ed9fc055d74f8c8d2492424e5"}
{"id": "train_073081", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "d5e56b4daddcd53415f83f61fbfb6126fcc5e5b8"}
{"id": "train_073082", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "05676b04bd270892109662a4555b65448b444eb9"}
{"id": "train_073083", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "8e3cecf94e3110d7a4e437f2264bcc20f86f52c2"}
{"id": "train_073084", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "1fe39dfc1492c7903742c4fcc99b2d219901e681"}
{"id": "train_073085", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a88cea903a1300c261010d3777de6d1e88e8a658"}
{"id": "train_073086", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "505b4fb94011b6df0a3b10a2b7d0c7783f8875ba"}
{"id": "train_073087", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "35d1276f3737bb3a75c8deb4b07cbd82bfe24566"}
{"id": "train_073088", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "1bb6393838b96d660042362a0c4198f2a9fc6311"}
{"id": "train_073089", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["performance", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "08d1ea526fbb5e6fdfbd1d02faf5ba0df45196c2"}
{"id": "train_073090", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4d31982b72de03bf8259228cb0f55b7bbd77d399"}
{"id": "train_073091", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "c3bfddf016b2c9f5c2633acbcc5121ce67f13864"}
{"id": "train_073092", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "df16832835a5b96adfc5a55e716d3f2e9effbe9e"}
{"id": "train_073093", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "5508b6c994bc717d804f6c7ad7e3bf9d0fd256b8"}
{"id": "train_073094", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "90cb27146f11f4cf4c0639916e935fa33e94ea93"}
{"id": "train_073095", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "c5a8c190d5bb177e147c27c8d05bce3ece7743ac"}
{"id": "train_073096", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["coding", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "63d448acaebe1f7a8a406914cce79fe541216d79"}
{"id": "train_073097", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "12e5bb9109bd900d149323062de445e73c87b444"}
{"id": "train_073098", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "9b681f0a2f1a2fe37616c5f1db54f41db378a12d"}
{"id": "train_073099", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "9faaea85eb91c55187d66811676ce4e9e391ea5f"}
{"id": "train_073100", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "0c5728ee4f8ecc8017e14d446a30db5114e7510b"}
{"id": "train_073101", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "8907c41f67ec18dd49292ab809eddf0bb0739887"}
{"id": "train_073102", "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\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "c7ecb6e626f862d6974051724b7c8eb741f0e069"}
{"id": "train_073103", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "a2a0ff054cff55eb05503366ba33b89479efa6af"}
{"id": "train_073104", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "dd50c5245af94ba7a2d6e0ea62ae58137d1bc12a"}
{"id": "train_073105", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "741196d184af649daf132e5fac753784a84b30b5"}
{"id": "train_073106", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "8bfd8b327497b413be18411898dc965dbb99d432"}
{"id": "train_073107", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "8410d6a036765298d7ff7d6e32feed8f22f6e0ea"}
{"id": "train_073108", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "75066fa7f3e25e093f5a2f9a50699701b2d995b9"}
{"id": "train_073109", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "fc8cfcfa36e6b3ef9f1c17e8d23b211568ba5016"}
{"id": "train_073110", "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\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "a9a3919a335b34fd80ef71340a3fc94269f731e3"}
{"id": "train_073111", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "8ea0d31afd0a7647dbef7423f01cdbf8d3d35a0a"}
{"id": "train_073112", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "50ef229b0bbdad1f0f760eec1981b1866aec73cc"}
{"id": "train_073113", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "a6abfb8f4225a3ac57f32bbda1a802280588ad90"}
{"id": "train_073114", "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\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "8b38da0b7fbedfc6948e90ce6cd35e535ace8d69"}
{"id": "train_073115", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "84346ffa5480c22779899a5e0d2514d1ea4bf885"}
{"id": "train_073116", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["research", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "083d8f0662b2468da0a3dfbeca43663a880f4352"}
{"id": "train_073117", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "b1a6560806f6079cf2b5f80597675214e86e06c0"}
{"id": "train_073118", "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\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "41b6269636a0664fe56889acea58db49e5ea7fc2"}
{"id": "train_073119", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "b86c5fe7d14bf6da58fff644c6d7586b51aeb22d"}
{"id": "train_073120", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "a344abd8592e10d753bf9298ed4d94ebc80f010c"}
{"id": "train_073121", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "cb7b2a68784395b1851dbbb3273b359eafdb4a85"}
{"id": "train_073122", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e6badc166a8c31d68aa9aae2bda6e39f1f71a1be"}
{"id": "train_073123", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "ebd0a44b1fdbf5fbba07d9ad0957049228d4e942"}
{"id": "train_073124", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "3f23d879a492914a5a3ec74e3b262b8be17fe652"}
{"id": "train_073125", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "1793d73fd443d51589cf6f63c1731efaa37b55bc"}
{"id": "train_073126", "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\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "5e178342408bd8f2dc9fe423d62f1304bce47245"}
{"id": "train_073127", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "89524bc5fd1d084e19660453ae870c49cb8ef09a"}
{"id": "train_073128", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a6d8065b732f2f3d8a42d290c84149eb04346e2b"}
{"id": "train_073129", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "d21734621c76bd3a72fb2a21fe654512b3d01397"}
{"id": "train_073130", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e771803bc4d0991ab8fdae7c88c3a52662508fdb"}
{"id": "train_073131", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "7dbdef615cb2e994d20af6d5591d3f953da0c56d"}
{"id": "train_073132", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "13b580eb16f6234c05cb186d2aa2f51ea07e5d6f"}
{"id": "train_073133", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "e5142bd4231b57b720bdf69a67c6b100744074b7"}
{"id": "train_073134", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "06fabb38ef8a49bd6e82253cc35381622a596262"}
{"id": "train_073135", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "a9be2e25a1c22e35ef5ce123d22b386a14d14769"}
{"id": "train_073136", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "fb967125194353c30b118e3f56e92cd1f2d8e74a"}
{"id": "train_073137", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "8ca780bd441c59244b5e19a875068946a8fe00dc"}
{"id": "train_073138", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "dd504db6f25e7ccde3cda45339006f0be05eaaa8"}
{"id": "train_073139", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["agent", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "10f85a2aebc8480559d989334d4129339de2996a"}
{"id": "train_073140", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "3ae1403d3751381ffb898850f9165189357e064d"}
{"id": "train_073141", "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\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "a515728dc09bf7e5a127f40dc769afb47f06f6a7"}
{"id": "train_073142", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["security", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "0d6ebbb5efa3519d12552de0d9d68be7fb93a587"}
{"id": "train_073143", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "9ca245306e3b4f99b42ce81da630672132d8d4f0"}
{"id": "train_073144", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "fe72144bcf5b844bb222fc31cf02744591a91358"}
{"id": "train_073145", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "e2b32dbc399152c7093cd34fbbe785ce215fe1cb"}
{"id": "train_073146", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "105dc8fd392145a4ed4b3ca52823ab90f8049a4d"}
{"id": "train_073147", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "678ae6324e32b2ea92e1005605a87384f785fa4e"}
{"id": "train_073148", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "38c81a1af994daca76648cc96bbe12915c80c920"}
{"id": "train_073149", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "2545ccf251a15b4fac1a9737916c9d9e2214308e"}
{"id": "train_073150", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f6640c7f53dc1dc6f245b272fc653aac3dfd71a4"}
{"id": "train_073151", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b0647451eeb3b4dc85d02ad687a3544682fa5d66"}
{"id": "train_073152", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "6a3241fc19a18814736c053d2fee58be00db08ef"}
{"id": "train_073153", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "ee575fed6e7e1ebceeb92059d12a65a866223d83"}
{"id": "train_073154", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "5243a52d1f9c186b5639608891b205ea7f9b7d45"}
{"id": "train_073155", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "9dd6c8b69e5d5d94438b88e288571a883ea818ab"}
{"id": "train_073156", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "94761ab214640dfac1099f1c6498edaaaaec7c08"}
{"id": "train_073157", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "b8e0c259ffa6dae9076806ed35691c3435fa2b7e"}
{"id": "train_073158", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "3ae2fc8d581c5b7437435e58aced22bed1db562c"}
{"id": "train_073159", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "481179026b261e82d9814873f9179bb36c8c291b"}
{"id": "train_073160", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "7cf34dda10f91500c236991bb97f66ce451d244d"}
{"id": "train_073161", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "bdfdd2bcb8b3ce897c0a5a3aec2f42ae0480aa30"}
{"id": "train_073162", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "755acde1d6418b614bd763c73679028d0ddf7c3a"}
{"id": "train_073163", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "de41d778f72686440cc318248ab64ad90ade3755"}
{"id": "train_073164", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "b3938080c93ef28d821c770f8b921dbf861c378f"}
{"id": "train_073165", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "ef4834b0b9fa79a46b7d4c18fa9e12ec11c89a70"}
{"id": "train_073166", "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\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c81b815f6f769fdc70d3494975076d7dc231daf5"}
{"id": "train_073167", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "9ee1be780ae7ce43a7c19007fb57f969f105ede4"}
{"id": "train_073168", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "b6ac84cf5ffc7c5e9e749eee9176217f5aae09de"}
{"id": "train_073169", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "61a73b41560fa2a27ce320254e6ce229b3d1c498"}
{"id": "train_073170", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["performance", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "4ed1cbdb0d1c9f80820c7df008c3e89dd3087ea1"}
{"id": "train_073171", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "25f814fd3f75629e613313c9ec608efe9ff14bb3"}
{"id": "train_073172", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "cd1a082ba9209c3062c487b687653932a106577c"}
{"id": "train_073173", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b16b2599ac3901d62b50c44b94bc174878a60f99"}
{"id": "train_073174", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "67d1264110dc921dc18d0440ea96609c0a799acd"}
{"id": "train_073175", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "bdb8f3e3a4f763b8d85ff5a0bd16ac1f622dd390"}
{"id": "train_073176", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "c30689d2a381538f70bfaa5ff16456360e9d1a25"}
{"id": "train_073177", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "456ece2f964cb0a258f106a77dc23ecde90d1e06"}
{"id": "train_073178", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "b17a8050cd5a9e9b475ec07535d654a2290c3bb8"}
{"id": "train_073179", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "280ca86d910435cd53b669318e39c3d34870e842"}
{"id": "train_073180", "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\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "02740d1df984a6decbef5f1adda62039dc6a5467"}
{"id": "train_073181", "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\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "4abeafcf269e788d65d56319f2d81fe9d0adced0"}
{"id": "train_073182", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "0d765b7f434abad08798530df51075ad6e0c1ae7"}
{"id": "train_073183", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "c393414786e250ff6da1ce4b400d73c73354c83b"}
{"id": "train_073184", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "fe7b6de905ffa2d36e8c28eaa95d8fe05a158b67"}
{"id": "train_073185", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "8f7a3628a73ea1b612eedac6265b80a87345290f"}
{"id": "train_073186", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "80754e7d9296271897091f6168b673a66e5b15d9"}
{"id": "train_073187", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f9145469b2597fe7a8d8fbed24a577e64b7c74b1"}
{"id": "train_073188", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "67a93f26145a46a0f4fbf526f65a928dd075e213"}
{"id": "train_073189", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "b102790df21105a5e3de024495945f1771c1eff8"}
{"id": "train_073190", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "8f1c8b521b7f511ba4d4caf3f856113a29571117"}
{"id": "train_073191", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "2a54b09531713831f187f2cc229a025049c4e65e"}
{"id": "train_073192", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "7dbb9c0a039753c98ebbc1d70a9e5b2d537bb693"}
{"id": "train_073193", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a28deee1c995f84bd06452add9c2cc46cc344f36"}
{"id": "train_073194", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "9913852c96e08c292095181b215c0eac41d38a96"}
{"id": "train_073195", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "10d60ed5943119f674357a37abe826dccbe30853"}
{"id": "train_073196", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "d03247656ae8b75754ba7ee42627f1916f2da571"}
{"id": "train_073197", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["performance", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "52c5f2273c5c6b3c29857afce1e964325f6ec6b2"}
{"id": "train_073198", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "00e9c6f144ea8d04b03ffd57f8ba86d5abe4a813"}
{"id": "train_073199", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "ca003ebce55b8150c68d2cac1ece8e4831dd16cc"}
{"id": "train_073200", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "8b6215ec57cd39c7df77018effbcbdcb2e4289c3"}
{"id": "train_073201", "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\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "d3b3d9d9e4bd2f27afbd2edf8228cafdd3aa9c38"}
{"id": "train_073202", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "1e30c84d9caa8e9bba1187c2cebe2e634c9a35c4"}
{"id": "train_073203", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "69a7735bba98017c670d75c5f9ee9d7be1de6a21"}
{"id": "train_073204", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "a26d367c085e0c10a7cfe4dcf015282567b1ee21"}
{"id": "train_073205", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "186e01f5f5b72b55bf59735daa70f3b1a707bce2"}
{"id": "train_073206", "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\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "2d223f07203e638dffe268cc45095c56255a823f"}
{"id": "train_073207", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "ea716d28335c4861bbcce4ec36b8f4a9ed99fe23"}
{"id": "train_073208", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "a0b8f54e6eda22f72d909e523b9a110e9e1abb36"}
{"id": "train_073209", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "7fac73e6c3cf048fe122c26a5e9eb91d99c83348"}
{"id": "train_073210", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "c68881973e87a6871fee85537788111f2c6f4abe"}
{"id": "train_073211", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "54cfaaff63b062c60c7d55126ca822a4f7c7409c"}
{"id": "train_073212", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "4cb281b23f79b2e1954d492cdcd66120391f2dca"}
{"id": "train_073213", "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\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "75003ff159127c1bdc4d0a0df221dd2a40d7c806"}
{"id": "train_073214", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "43b52982b67ac44cb47daefe6abca95a1eb5a4d4"}
{"id": "train_073215", "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\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "10061fe6fc31177307aeca0f0bef2020192a595d"}
{"id": "train_073216", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f48e35c77c72e2cdea9734b8d5aa3384b3eb2429"}
{"id": "train_073217", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e089ea2c453f800a7e2a738b36421fa22051b576"}
{"id": "train_073218", "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\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "7a27415268eff5952b27ec375efe60a94f134ce2"}
{"id": "train_073219", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "38f8337fa510ff8cc76a64be69c5914b7a5bead5"}
{"id": "train_073220", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "8881b11d6dccd4310274e828ee7af9bc53696952"}
{"id": "train_073221", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "f316bb570498739fb286c4553cc0b23afafbb4ce"}
{"id": "train_073222", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "cc523962d8c96da221512aaaabeb9ab8036b9a95"}
{"id": "train_073223", "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\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "32eae54258de61f8a51795c2acf742934995d853"}
{"id": "train_073224", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "d97250aee671c68ee2e265d97e7abe4cc5b5cb85"}
{"id": "train_073225", "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\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "14228746434456c3e54883cad34d17e834c54afc"}
{"id": "train_073226", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "9daf6b1b752261710fe494ab075365d7f39d954a"}
{"id": "train_073227", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["agent", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "0d0052e71974e593b1654ccf921edadf87ab8cf3"}
{"id": "train_073228", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a47673ca5186e1067c999b7d49df71fc2dda1dc8"}
{"id": "train_073229", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["research", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "65ffbf2208afe9e075075d7b33fc7bb504d85644"}
{"id": "train_073230", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "cffed426d93cd0c9de5bbfd78b564009b2cad21c"}
{"id": "train_073231", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "1abd918b653e37231f793ec9c1bbfd0980816b90"}
{"id": "train_073232", "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\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "87aafed8f3d8adf2408f9057ff4ca39aa1a947ca"}
{"id": "train_073233", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "0925d32df13c8139447fbd569dc6c7cbe110db12"}
{"id": "train_073234", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["research", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c930d1e8383e7b4f1d51501e223fafd88c7521b3"}
{"id": "train_073235", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "19ad38db1e1666b4fded635224998dfa5afe999f"}
{"id": "train_073236", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a9fee9abd5ea1e1ec82dd11b503b271c66e8a5e0"}
{"id": "train_073237", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "bf54c5a88bda7485bf25359740107ce11d308478"}
{"id": "train_073238", "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\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "910d60007108eab9f9716ae03580f023ff2f06cb"}
{"id": "train_073239", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "b12091eb2ac6ecffb5e49b8711ba0df321a2763d"}
{"id": "train_073240", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "36a16d43d511a8e7888b6012c231e7b2c15359c6"}
{"id": "train_073241", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "d23e3dfc7c65df4168e941a145a87e4afe9484f2"}
{"id": "train_073242", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "b88e27a3d2e44a7bd7d97e381f438687578dbef7"}
{"id": "train_073243", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "041877e2e0696f5deb5d08cbe56fe4b04d5aa947"}
{"id": "train_073244", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "4fa77f57082769092593e5780d27dae462a0502c"}
{"id": "train_073245", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a665c7853ea1cb16bf5785721d9723e81341c9c3"}
{"id": "train_073246", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "cb7628838ad56508eabb38bb56cf23cbac19d9ce"}
{"id": "train_073247", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4a753a78f2ac46ceb5a857ca6cc6032cf36d0017"}
{"id": "train_073248", "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\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "24ac244f10f6540513ad310b63274b74ba847ad0"}
{"id": "train_073249", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "25baceac7043f47728e733e4e8a6d1e17876d7e1"}
{"id": "train_073250", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "dafccc12cfa4c7e8c554a6d95261289cfd4c153e"}
{"id": "train_073251", "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\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "feb14c31ff7538927ffd10f342a7d57f7bae9dd4"}
{"id": "train_073252", "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\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "b5ae0d40c35e307879e11e2d3d6c2840ce7cec2f"}
{"id": "train_073253", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "34201a250b9c4de3b5c2273b3e167c3b1fe63c6c"}
{"id": "train_073254", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "66abe77e20d588342018e75fdab3544d38411a78"}
{"id": "train_073255", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "6b8c6efe6936be64111a63deae6a0b664c67cd75"}
{"id": "train_073256", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "ce4676e413b6e7226ae913b281fcb51b6fc05315"}
{"id": "train_073257", "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\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "7a691169fb6932c58114e7ae34d404451743da59"}
{"id": "train_073258", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "0edf02213cb2f8e6382ef7ca720ee722a094c968"}
{"id": "train_073259", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "e98fd2a249a903a93d924659ef7aaf4f1c9d7fbc"}
{"id": "train_073260", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "8180bde714b32f871e04342d4023198a8cf20a09"}
{"id": "train_073261", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "a776c6c813064fdb075bcc9f209e13b871cd795b"}
{"id": "train_073262", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e3a8dfe81a22afbcb2edb1f43a739b32a211622c"}
{"id": "train_073263", "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\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "75f205a378bfdf3c2d2177e4fd56fd74476c0e46"}
{"id": "train_073264", "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\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "5536268005bfd71db011413c93558a134661e8de"}
{"id": "train_073265", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "9e750cf447fae0335f50ede631a1002c5697b215"}
{"id": "train_073266", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "36a16d43d511a8e7888b6012c231e7b2c15359c6"}
{"id": "train_073267", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "ce8f31e33dd3771f2bd9a9830e818894bbe21845"}
{"id": "train_073268", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "3aa9b2b6ece9869be258828e0490b432273920bb"}
{"id": "train_073269", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "462649ccf5bbfeb4cba164bd59ead150ec362603"}
{"id": "train_073270", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "61edd534c79b209f64c3c6d96a44814b1eb09140"}
{"id": "train_073271", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "23b2ce3f615c23d2d0394e30903f2532792eb354"}
{"id": "train_073272", "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\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "4bcf63261060e19f846f4bf0f32a1c6199068fd7"}
{"id": "train_073273", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "8664751b738c71a395e8b2927a9cfa4b3890ba3f"}
{"id": "train_073274", "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\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "5f14db28b948f90f67f06e80682c1bbd45713077"}
{"id": "train_073275", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "47dbfcd8f185dcf74d55cd5af473b5d442b63a67"}
{"id": "train_073276", "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\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "4e5c3f4b46e0dec070e0983099c609f392f12194"}
{"id": "train_073277", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "9e69cedf0d37ab64439ec5205d62f8eb1442b056"}
{"id": "train_073278", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "980fdafc0e3729b5d655712fc2ac4735f58f8676"}
{"id": "train_073279", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "f6e831c06218a11d6cb6f2f378cca465cdffc885"}
{"id": "train_073280", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "7893981706ef50278b1f01e2fae94964b429a068"}
{"id": "train_073281", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "128b2fbbc27854eaa097ce84a44ea0a5ada53c11"}
{"id": "train_073282", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "72682fb9057eb9cd6615c465502eeb613810710c"}
{"id": "train_073283", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "2ce7a71e98b89a91328614a3783d6ad154c8c27b"}
{"id": "train_073284", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "00f6b72a17fd711dd23ec429e0ae5244348400e4"}
{"id": "train_073285", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "e3f695704062f8a1aa6cf2e77454ea5200eb3dff"}
{"id": "train_073286", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f772af691937f0455bde968372aca1078cdc65cc"}
{"id": "train_073287", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "9774d9be424034110b6f4eb10f5903c46f493c3d"}
{"id": "train_073288", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "79a68ac9c237972cac9a15ef2093ac3ad5cb6c23"}
{"id": "train_073289", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "b0f14b6e73b3f7fdd33d93e9c5442f0eb48b3692"}
{"id": "train_073290", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "2a9d4835f9989093bc6ea0c82fb6c142003ff6d3"}
{"id": "train_073291", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "2b59c3068a99fdfa99bd6a0b96e4da2e651990a7"}
{"id": "train_073292", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "1c110a02328f609e8bf7f1e2840d2863bba45efb"}
{"id": "train_073293", "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\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "1e056b810dce42cd27b4aa64380307ec9ff2960d"}
{"id": "train_073294", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "65ffbf2208afe9e075075d7b33fc7bb504d85644"}
{"id": "train_073295", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "1ffc0b99889a2cbe808fd25582995e68b5ba2b7c"}
{"id": "train_073296", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "01ab69ba6b084f4a5165c969192a6f8087abbf36"}
{"id": "train_073297", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "e0fa594653de5e8a7a795216901c2cc64b7bd2b4"}
{"id": "train_073298", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "bdc495332ef7637bec2de9c382a7c9261f376c30"}
{"id": "train_073299", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "9f0b9dfcf39d3d780ac229bdb94fa7545cdf3875"}
{"id": "train_073300", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "cfcdce3ddfa30a182536e4065349c3e1870619bc"}
{"id": "train_073301", "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\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a7b319ca448f4cf5e1d4e2d7816d29a7542d2554"}
{"id": "train_073302", "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\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "8657a6d485e0b3f1a5ac7a8f7d8016aadcfc802c"}
{"id": "train_073303", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "ca563dca259269439e429be34c5800d9f88c3faf"}
{"id": "train_073304", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "373f2ec7c031381892f74d7f73eb19f1e029e4b0"}
{"id": "train_073305", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "9407b22252b0f126355554733d15db7936d0c99e"}
{"id": "train_073306", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "6b1be8834843489802aa8b496b37f3d9442bf7b4"}
{"id": "train_073307", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "75066fa7f3e25e093f5a2f9a50699701b2d995b9"}
{"id": "train_073308", "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\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "6173472dff30c904adb5c0d4f1de47630c502ab7"}
{"id": "train_073309", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "2eb8f2fd383ec0110dd1bf0474a65fcf8d518a78"}
{"id": "train_073310", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "4a5382b8f644ae6c75c6c6fc1a5343342a7c76ef"}
{"id": "train_073311", "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\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "1edbeb384647270afd0e8fb392cc5e6d76d35857"}
{"id": "train_073312", "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\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "0317b8417a2e623cbeeea9512d5d44aa5bf52574"}
{"id": "train_073313", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "b0dbab00ffbb7cf157f61d33898d6b74d5099705"}
{"id": "train_073314", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "e22d0ee9ec118dbabfc1881aa2e9a5550aa7b791"}
{"id": "train_073315", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f25101a2a6cf7c32177dcf6ed0c7e164115b3eea"}
{"id": "train_073316", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "6b756ecdadccfd91a4b1d462a222ecf42d8cf358"}
{"id": "train_073317", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "2eae75e7b4edba4ea9c09e389e71f0bc48df1eeb"}
{"id": "train_073318", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "7bdeb420ea6dd790e74a38a8ba903f93e25269b6"}
{"id": "train_073319", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "8cc254f8ef077196953ec11f06991c2e18ddc5f6"}
{"id": "train_073320", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["agent", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "91125ae2022236376df9a92a10c927b79b6379be"}
{"id": "train_073321", "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\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "fe56adc6b1bcfe520a10d99e85f9e373bff3362d"}
{"id": "train_073322", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a6db63fdaef3da80ecdb4e1142c4457fec173f98"}
{"id": "train_073323", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "edc5d03752ac7c0fde8e2bc5f484d40032aa4c0d"}
{"id": "train_073324", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "fc60df4ba04f526d8e2dadd22d5fc8161c6645fb"}
{"id": "train_073325", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["research", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "c22ad4b2009a9d9b0d45c2d9c8a9a720ceb0f1d6"}
{"id": "train_073326", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "c5402d7171c3940f78a8a76152079750e202ecc4"}
{"id": "train_073327", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "051df7b5a82096e64b216158f0220a48df2afba4"}
{"id": "train_073328", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "9179d007fad0cd04556bf5cb51b3848f045c3308"}
{"id": "train_073329", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "ecf68be4913f3de8b49f6add502a2e3dc4802d64"}
{"id": "train_073330", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4f1817cc1be69b7662fdf1de4da125570568dd80"}
{"id": "train_073331", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "934ceb158c877baf68d2be4e3e4c09c1ed8c4dfd"}
{"id": "train_073332", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "1893ce34b2bf16d2e8b128843001e230e4c91382"}
{"id": "train_073333", "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\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "eef79851d0679a571745af44232d740f72f9804e"}
{"id": "train_073334", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "184167d197ca05666bd3a8a42e2a7437fe37d1d3"}
{"id": "train_073335", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "3f92c619430eb179a284791a5f5141b2ffdc732a"}
{"id": "train_073336", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "e09bcf3021369e934c2383f6084b9061e8971fa6"}
{"id": "train_073337", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "16a8e4b4e7bb399bcaaa40113de6a8e297f8c493"}
{"id": "train_073338", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "2535b5869042407e38e9f1fe79409b5d8e72ef29"}
{"id": "train_073339", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "2374bcc2a412bbe2388334e67435280a45028821"}
{"id": "train_073340", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "980fdafc0e3729b5d655712fc2ac4735f58f8676"}
{"id": "train_073341", "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\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "4dd9d29af9db9802a6f57a40d8fcea1c7ca431eb"}
{"id": "train_073342", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "5876c27b8b8384fd7539870f955c9c67a5c6f2c1"}
{"id": "train_073343", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "5587280b315ba3a53d220f95a4ce7da21d9143fb"}
{"id": "train_073344", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "d16f5c296d6f8c6c0546273313d60755ca8ad6c4"}
{"id": "train_073345", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "81efcd19c23f4447c581b1a1b07934cdd231cd46"}
{"id": "train_073346", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "8d859ad647fb2010c54b58d104d0c027280e0324"}
{"id": "train_073347", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["data", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f2b6bcf6d33c080496d5d750a72e8cd581baa786"}
{"id": "train_073348", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "6371779a5a5ab76deecfe0f375ca19b64885d855"}
{"id": "train_073349", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "1a6dcbd505b180c3084b3f003677e8a660f6aaa7"}
{"id": "train_073350", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "e4bcb9da36fa600936f7c8755009752359f01282"}
{"id": "train_073351", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "28f3c9efad2040432f27573dd3b2b8e0a899692b"}
{"id": "train_073352", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "cf3c877eba8979262da591a7e4bae64bf7a74d9c"}
{"id": "train_073353", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "5b4a42a3ff5f263df4ad375976a6c5b52cd42524"}
{"id": "train_073354", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "a9ff2a9348def0cb305109a2cb3112ff12e20f94"}
{"id": "train_073355", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e7120e84ee797756972c1057a33b67940b2cc9f3"}
{"id": "train_073356", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "c973c1b4bbe813839b702164c5638a1cb9c065aa"}
{"id": "train_073357", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "240462aaf8fe71b56565a10386e39eea42511e19"}
{"id": "train_073358", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "033d022fa0a879bf47b6401e1eb9fa68035b11cd"}
{"id": "train_073359", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "4c001b963af9ecde9d36065917db39e0ef7e5375"}
{"id": "train_073360", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "11b24a044764b055eb637d92deb5d6b7f197e9f9"}
{"id": "train_073361", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f4e1667b355292ea2e02dd21e89fee3a7742abf5"}
{"id": "train_073362", "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\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["security", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "0ef683122b4d0b0f5aa3d683b3a625ac6f9e7c59"}
{"id": "train_073363", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "872a8af2daae1185d36a81d62186c6c4c9da1314"}
{"id": "train_073364", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "b1ac1fc910a3b4f3a98722b55214550c16d42710"}
{"id": "train_073365", "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\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "fb9ca4d5b9fe8574655ea304f0263fca5af65c7c"}
{"id": "train_073366", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "ac018b34fdce58c50671e2014e9fbeb0cbb4c364"}
{"id": "train_073367", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "914fe433a13c4e65c0a2747a7377c4d498711799"}
{"id": "train_073368", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "a8ec7e2688508047a021d98c4284576f3a307e66"}
{"id": "train_073369", "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\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "06880ac1e401e382ef7af054699c57c988e0d0a5"}
{"id": "train_073370", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "984f7ec4db37387d91625aad15c956be5ca4143c"}
{"id": "train_073371", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "f867d366f34fb11b2df73fe7ff6a95b775347c4a"}
{"id": "train_073372", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "61fc020b2f0a24b646886a8ee4ce525e7275d23c"}
{"id": "train_073373", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "16f24ff1a13c787a9a48ca5b122178fc33ea5d72"}
{"id": "train_073374", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "af44a8f3e16d1706ec58faedfd0f558f9d576e33"}
{"id": "train_073375", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "a626003ad6efee3efc2b53034c134b50390c8fae"}
{"id": "train_073376", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "6494a422452f104f3365c02debc942ada60eb0a7"}
{"id": "train_073377", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "6ab66490760f6f9d3b88edd5b8b25d55b38f4d5d"}
{"id": "train_073378", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["research", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "1c564cb6aa5627d5000c06df42b09dee548b4d85"}
{"id": "train_073379", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "88207863e2bae6bdfd00cd03165f3bbfd2a5c34a"}
{"id": "train_073380", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "512dd00b6c21f44c4ea77d5750d3ecca526da5cd"}
{"id": "train_073381", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "aaaa28bfcb6343a7f820b6de26f8c408c517739d"}
{"id": "train_073382", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "cc5fe6df6d67fabc7c65e482a8ec8c1b8c5a44bf"}
{"id": "train_073383", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "136bb8fcc51a9a6e3f38ef0f7950fc31e518de55"}
{"id": "train_073384", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "831a5a3e46cdfd5a44a7615ff3bb26e4a2c9d9b4"}
{"id": "train_073385", "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\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "5b3d597eb88cf9320f7de377351e0f4481f54149"}
{"id": "train_073386", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "11155d94d982a1eec97b30c21a7a87350ca9602b"}
{"id": "train_073387", "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\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "c62455d25b7b30b7d15d56ff66f0b97cd5ffe2ef"}
{"id": "train_073388", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "016ee965bf128ab58386bd98e14d6761f58de5e7"}
{"id": "train_073389", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "fd4c6afbc6c48b854697e06e8be1d75f013701c3"}
{"id": "train_073390", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "a418d861e1217bff3756415c1c070824ec288fbd"}
{"id": "train_073391", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "35cd2fa791fb3a2b46a3bd9bdac6c9fb4c9825b7"}
{"id": "train_073392", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "05676b04bd270892109662a4555b65448b444eb9"}
{"id": "train_073393", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a1e91fb57456720621b799cc80b6c4ef58784f47"}
{"id": "train_073394", "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\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "0de7d928d243dcb27ff070283ac3a520f8d8e91d"}
{"id": "train_073395", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "7b29174a820c3ec88472d129e8a24d173170aae0"}
{"id": "train_073396", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "1405ac8baf6dcfb87b75620d0054a48febf33131"}
{"id": "train_073397", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "d45bb2d5ba4dd3eec842589cdfe0cf7fa9dc7747"}
{"id": "train_073398", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "012d3d1a9dc582078335444b63d253d507c4c54d"}
{"id": "train_073399", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "f230463b196dfc7a29234a13cf4d0c2fd742b75a"}
{"id": "train_073400", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "5e526437c3773e7d02cb829d299940f82948bd64"}
{"id": "train_073401", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "2af146be4e68ef71b815eeb735f7bd8858a74511"}
{"id": "train_073402", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "cc2acf8b8e7949dabaff5258fdd3394449834fb3"}
{"id": "train_073403", "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\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "e7ea49b24165988af212cb486aeb7dc6f1d5e635"}
{"id": "train_073404", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "0822fa919f9e77da687d6da27fc3dc3b0bc585d7"}
{"id": "train_073405", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "dbe71ac248636d3a50cc2fd495545bff46d58e94"}
{"id": "train_073406", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "31187d3400db45239815f1528602779f018c331a"}
{"id": "train_073407", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "4e861a3518a29cf86d8fa457fbcb4fd0404e0c74"}
{"id": "train_073408", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "3c3b89e9be57a31d42372f0a9996afe792bd765d"}
{"id": "train_073409", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "bfda0c735509329ee19f2c10f0a3dd317aa6fe1c"}
{"id": "train_073410", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "6f02359e139ef17e66177b22350b8332c27c2396"}
{"id": "train_073411", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "767bf58ebd0b1c3ce108c0892d5ed79229334c17"}
{"id": "train_073412", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "4a6e9726e1979c66c6ab4cd6d5c40e30e95ca6a6"}
{"id": "train_073413", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "3249996b9df0448c36026ee17ee620e49b2e042e"}
{"id": "train_073414", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "ae533309dc1bbed3658f0c06c66902619bf0bca3"}
{"id": "train_073415", "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\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "b2fbf978e8c01760f6dd0946bf1cbd7a5be9bb2c"}
{"id": "train_073416", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "28c219244247278d4aa622280d4a3e77588ca4e9"}
{"id": "train_073417", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "689e9eaaa3c65ae094f0f7def4fea51451f0a901"}
{"id": "train_073418", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "4d630552eb4eb359de5dcc7f8ae1bd4bfa353a23"}
{"id": "train_073419", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "54d2f174e43748eae94bd2244cacaac0172e58f7"}
{"id": "train_073420", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "1dda736b886fd752674bb20da514f0c73648d340"}
{"id": "train_073421", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "4587a096d08323d36d7a9ffb019bffd981de7d00"}
{"id": "train_073422", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "26dbafae46cf979bf1d27fed5033c9a7b0dc8e7b"}
{"id": "train_073423", "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\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "023e27c45b3c43bacfb0f40def91b8c1ad683b94"}
{"id": "train_073424", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "63ddab33410cd39fc2df9d71fbfbd483bb8a452c"}
{"id": "train_073425", "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\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a5d1817b7f84ecc034e39edf20376391bf28a641"}
{"id": "train_073426", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "75d397f4d0c1a786c8cf120b52382f29ed7f205b"}
{"id": "train_073427", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "3aa6475f611ec4f8628d598d6789f8986546559a"}
{"id": "train_073428", "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\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b2d79f1601d08ba455c0d76d285a90c904e8ae4e"}
{"id": "train_073429", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "acc12ec69f7f6e28b6056bcf1f94f7bf2d649a70"}
{"id": "train_073430", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "d7c3a2582823794d3e21938ac2861268c8a9855d"}
{"id": "train_073431", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "c6348fdddade64bf80b9a81a58c25911d328912a"}
{"id": "train_073432", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "6b470c9ea1316e0929d24ee7f9b83554f50fcdd4"}
{"id": "train_073433", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "1e30c1aa245bdd5269f426f8b49fe2146969283b"}
{"id": "train_073434", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "438c374bc877ef2b6fa54842767e2cb6ae1ab5fe"}
{"id": "train_073435", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "ce5d647135b99d86136a9c57c3c9aa1510af2052"}
{"id": "train_073436", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "0fb308893de6eff21775960b3173a9e59f5429b2"}
{"id": "train_073437", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "f4e290e1cdce44c4bff8efcf81b0851fd2055a80"}
{"id": "train_073438", "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\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "9d675e65bf253363f186a93d588272a404e37fbb"}
{"id": "train_073439", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "47ae8f2f6dca3c26e96d7d951cedf982e08129b6"}
{"id": "train_073440", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["performance", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "83a6741851b950c4a4de43a581dd6fffa10faee4"}
{"id": "train_073441", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "b16b2599ac3901d62b50c44b94bc174878a60f99"}
{"id": "train_073442", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "a998bc8b814cffc83331a370f4e3dc3ecab7ccb2"}
{"id": "train_073443", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "eaf09b41539299b2db2d8353f22670b6fd5af487"}
{"id": "train_073444", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "cb05d564f8b2a58cffb61f1c122c0776c2720946"}
{"id": "train_073445", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "1bb6393838b96d660042362a0c4198f2a9fc6311"}
{"id": "train_073446", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "44500a241d01ed4b386f2ff8948d7f670f36cf66"}
{"id": "train_073447", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "9faaea85eb91c55187d66811676ce4e9e391ea5f"}
{"id": "train_073448", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "c358d731ac518f01cba27e5dc2cc8e392ab94ef6"}
{"id": "train_073449", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "b34010549b01d3bbf7af0b8a4f4d32495e6dfeaf"}
{"id": "train_073450", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "b28744a4c20fe7a8aae3a1752456827ec0acc1b8"}
{"id": "train_073451", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "d1a2922e625b788bb94c56d26466fd2bcf2336eb"}
{"id": "train_073452", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "54bf111be7918eb841af2b7b047da3f3ef6f4e41"}
{"id": "train_073453", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "f5d0fd588dcfd2bcab65e4b8c474e419d21d0242"}
{"id": "train_073454", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "44f81d9c08f07b359c3dfe17d66a253aaa0bf331"}
{"id": "train_073455", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["coding", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "5c515f7cf9211171a1d24df23c981b2b198b2d4b"}
{"id": "train_073456", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "b17d4a30c8d92706d367a3d86c276265fa6806c5"}
{"id": "train_073457", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["governance", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "07b70a50b3e353b0c868855963c46351c802e887"}
{"id": "train_073458", "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\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "55be245127c25ae089d367edb076174319d7e76c"}
{"id": "train_073459", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "934c99875848e8f328aa75c0a32fce692821f863"}
{"id": "train_073460", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "f0998af0acfa5ba348f19b1539aaa8ae35479611"}
{"id": "train_073461", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c02697e8aec2a71937d2f383f4b7bd97997479de"}
{"id": "train_073462", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "6b76377544b993163245f6e3b319361b6cced008"}
{"id": "train_073463", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "ac5ed9855aa4d88d8d48c3b79c727ac48f224caf"}
{"id": "train_073464", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "641cff9996fb4ab584b8b60c7aec518e746ed5be"}
{"id": "train_073465", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "0f0d951bd382cf4f9c9e1d28ac4e115392aea05d"}
{"id": "train_073466", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "4ecbaa478176ecdfbed65ca15236e84017ec5678"}
{"id": "train_073467", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "3bc5b9cfe34ec492a5c4fcf6dd5ed79ec2fe1437"}
{"id": "train_073468", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "f1ef284cc136beba9e639469bf2a95ee21a30e3b"}
{"id": "train_073469", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "5c7fa5c83030db25736f19e892c794d369a4a3ea"}
{"id": "train_073470", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "b320f1a9bbbb0607c1561cdc3dbd5283ceae93f9"}
{"id": "train_073471", "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\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "7fa828196459030882d603c73545915355507230"}
{"id": "train_073472", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "61f5b952830451d53ea49cacaaea99c7ecb9721d"}
{"id": "train_073473", "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\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "7fdd44fed47989e8c45b3c0e7259b378bdca27b3"}
{"id": "train_073474", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "2791b2070ea720804d05047c56df9e8e6a8645d3"}
{"id": "train_073475", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "184ba9fad9b1029390b61bc67aa92036137c0f3d"}
{"id": "train_073476", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "d48b03488fdf0fc3791ce009042b227d706bc66d"}
{"id": "train_073477", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f7e85c736bbda55208f564a227a99433ce48cd42"}
{"id": "train_073478", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "f1385edae21e937906dc71629e6056d9bcfd4e71"}
{"id": "train_073479", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "6073ee9db4be70b2b39bc05b17b22dafff13938b"}
{"id": "train_073480", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "06365a9adaf46dafa457b286eadc9d2e5a12cf57"}
{"id": "train_073481", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "9dfd330b5595c10aeb933e826a1a0c753a29f593"}
{"id": "train_073482", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f45b7cad15b7e949df17c0e9bcfc46da013bbdd4"}
{"id": "train_073483", "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\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "56cd19642a5b83690b045a5890b7d4f139d6ef47"}
{"id": "train_073484", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "924aa5d71ec34035da3bd59eebfd1b6f05ce3b7c"}
{"id": "train_073485", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f39f0a2b80a51a3e4d63145da2ba738b191a66b1"}
{"id": "train_073486", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "aeaf616d20c95e1aea1ac2a376289071e2ba4eb0"}
{"id": "train_073487", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "21fc539c403a6e009b1882cc8b2f49a16ab00bca"}
{"id": "train_073488", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "fdfe92dbcd6b29679a5a5aa9830643d9262e79e8"}
{"id": "train_073489", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "30391bae5cf05ce3e1841601bfdcbb59be69e853"}
{"id": "train_073490", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "6500953aaf26a9f82f2e6909ded1ae85da83a3ba"}
{"id": "train_073491", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "ab32751f31abca15683ed1022c5592f043dbec05"}
{"id": "train_073492", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "15fcc40fffb56450f794f3dcf43a9ea486cebe8b"}
{"id": "train_073493", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "0759c1b247355b128cf80108f90c775700e968c9"}
{"id": "train_073494", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "52c58f89bb4a286c4f36b126bcbf3027bb01dac6"}
{"id": "train_073495", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "41b9e632178b5b5070705a2b39d6ead97a194fe8"}
{"id": "train_073496", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "d17be0472afe0c6f46655fda8bcb978f88c4e9fd"}
{"id": "train_073497", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "a7b77d71bcf7ac0269e6053d7da27f07d0fced93"}
{"id": "train_073498", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "9248cecaabe175264a0236f459d4214445bb1b8c"}
{"id": "train_073499", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e75c44d488acc09849e734418c967255f9ea92df"}
{"id": "train_073500", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "aadaf342210c5f5289f6f5cfece18660d1f0cf60"}
{"id": "train_073501", "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\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "ca498854a5f5c45d971b28c1a2c963a86c2660cc"}
{"id": "train_073502", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "7e542f7bd8e1baa3e961b08f4887384ae3b1db93"}
{"id": "train_073503", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "3d364663fa34b7603dee80e7b3c58c6ea21c1b57"}
{"id": "train_073504", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "e09169b21a81bd641f47799edbd8c75d7adde409"}
{"id": "train_073505", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "8e36e4c9ca529a3e667dc605dfd88c8ab0081802"}
{"id": "train_073506", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "60f5906cef2bc006e32fcf38458e8cf497ee8d01"}
{"id": "train_073507", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "027921a55bba5b1543ab3b3b844e74340b5e9be0"}
{"id": "train_073508", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "fa947277d0603b74a1c127a5598d674882ffe7e7"}
{"id": "train_073509", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "ed04079bccc150641a86a5dbaa82ffc35fdb6530"}
{"id": "train_073510", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a8d8236315cc49b4cf29765c4664b7526aec1528"}
{"id": "train_073511", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "3bc5b9cfe34ec492a5c4fcf6dd5ed79ec2fe1437"}
{"id": "train_073512", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "a673fc3910f348c4481ce8f164949dd95679ffd7"}
{"id": "train_073513", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "dac787b76c41d05486af81903dcab00658372008"}
{"id": "train_073514", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "6457ae641b66ed2bca0fc03321b007a51e59797f"}
{"id": "train_073515", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "2a10bc6a89c998d3282bdc56d110ad0b4bcf1781"}
{"id": "train_073516", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "da2aa776c5e3ff08a98e07be80d34030c8073d4f"}
{"id": "train_073517", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "23e91dd97ca412201230ece69adf651c83c29682"}
{"id": "train_073518", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "ee57da888fa514ce40d7770329d700cdb222be18"}
{"id": "train_073519", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "2f17f4ff54a8db76864c4f516a77f4778f58fb49"}
{"id": "train_073520", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "b5aed132fc7a9dab81cdc3bd1b7d0ab1ecb71905"}
{"id": "train_073521", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "f7a9c579b586c56f1fbe640c0d8835a5164f31b5"}
{"id": "train_073522", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "dcfdece8466f01bd6a94d90c25c889a98f5cff1b"}
{"id": "train_073523", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["agent", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "4e0025445fd951a8b286447ef652b7e53d309e2f"}
{"id": "train_073524", "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\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "ce07c98a5d9e69f65a5f54c4468647bf066cef67"}
{"id": "train_073525", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "4e1f5821fc3d5dc19bb148e82b529072855ace33"}
{"id": "train_073526", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "739ace49a7859b2b65591baf9f50723e53e6b2fe"}
{"id": "train_073527", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "942a43b8cfab0645b0220c6a5bb228f22555a5b7"}
{"id": "train_073528", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "701ad2c629f555aa0f0b5927158d856c96ae3109"}
{"id": "train_073529", "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\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "526bc0fdd979102f4bfcf8dc8bf545def6244e1e"}
{"id": "train_073530", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "cdaddb9931dc6581a22fbf1267be3099d1224630"}
{"id": "train_073531", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "cc24798f77ff93deffbf1cd6658435e7fad2b16d"}
{"id": "train_073532", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["security", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "b43a11148e936fd7d0bf13d5303ad9493818186c"}
{"id": "train_073533", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["performance", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "395c86c75a33dc67a997c9284acd5723f6113d22"}
{"id": "train_073534", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "98e03c8ce74516e3f15e7be6f5b5325295efa3be"}
{"id": "train_073535", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "508a68a32c018907ef856929c644577a8485b631"}
{"id": "train_073536", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "0893e2beeb08a9785ee78602535b6e018921bb0f"}
{"id": "train_073537", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "fe2bd2e1ba29d6e619955de07be50574c5f5962f"}
{"id": "train_073538", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "2e18b6e481b91d07f584b54cacaaf281179c1743"}
{"id": "train_073539", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "52aa015f1bc1ac486c093967c0501553a448bb9b"}
{"id": "train_073540", "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\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "ff6ead12fdc96534d2c6c91c543d73b6e6fc5659"}
{"id": "train_073541", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "83c01b421afb3e5afd7c91da8317242ce3aac573"}
{"id": "train_073542", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "6538a4711c511898af65193f06810b2dc426ff22"}
{"id": "train_073543", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "ed72ae003db6d1203c9757fd56860031a4741a80"}
{"id": "train_073544", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "03cfabc56e81faa5c08e76901c58f7312dc2d5d4"}
{"id": "train_073545", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "c44d2a3609e7b3b27ee3e6cd96caa76f975a8136"}
{"id": "train_073546", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["coding", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "38ba6f9455b9c3d77f901fca61c2eb39237e7b01"}
{"id": "train_073547", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "c228268d55ae894870754688b3293a2ffb5f4b55"}
{"id": "train_073548", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["coding", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "88ee0be9f04a96f39087d932894f686bda65ac29"}
{"id": "train_073549", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "860be4a8ae9ea0aa1af6c37c7d30abafaf821a26"}
{"id": "train_073550", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "3ac01de5913705a54369fa4d192383e2b2f2c9d5"}
{"id": "train_073551", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "18e368b34ec4300a37513ab8fc59ffa53686bca2"}
{"id": "train_073552", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "1ce81d5d76c6eae389a61249ff13453bed298af2"}
{"id": "train_073553", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "19effca63d3dd31129b8720f54134e5b8cb55715"}
{"id": "train_073554", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "b3da9aacd6d0d36bd99dc9aee8c212c19894e902"}
{"id": "train_073555", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "9c7ee9a176c4f0c477928e6592500935cf0795dd"}
{"id": "train_073556", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "f32c57008334add28b00b21641f89d2ab34b9206"}
{"id": "train_073557", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "86b54121030b3c09e9178ffe14808619db8496c9"}
{"id": "train_073558", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "2916f52e264c0b5c32a34c38d3444be0f0a1a654"}
{"id": "train_073559", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "6fcbb3784a0625492453347d0a7c2093968d4783"}
{"id": "train_073560", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "f990c3e314da635b2ba55a8fe67188ba3db194e8"}
{"id": "train_073561", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["performance", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "857ec5be3c888f4204ebc2bc50e1fec3e73fca8b"}
{"id": "train_073562", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "495009c58cac154858dbd4677015c4f9736da708"}
{"id": "train_073563", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b3ff82af07a53780ca1509452ff5d68bbe96e002"}
{"id": "train_073564", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "ac224a2cb01fc1bc14f44dad6a2f6022d1015e8b"}
{"id": "train_073565", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "5e627d63c08b0da79ed677d7d7097f8b51517cc3"}
{"id": "train_073566", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "4d2776efd72e12a145c5a382e9eaee610cbbc2a8"}
{"id": "train_073567", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "62863bd1ee2c6882825c04056bcecadd8f53bfc3"}
{"id": "train_073568", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "a9c07c44829250f21152d5939ef3bd27dd8549f9"}
{"id": "train_073569", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "be956a3b844d6264cf1d02549847f825cf1fed2a"}
{"id": "train_073570", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "2d17dfb27c39f75d5cc5e5c99034c5490f91b02e"}
{"id": "train_073571", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "6b439307e49113b836e72d1429f50837336d0d76"}
{"id": "train_073572", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "333396aae072ba7439b1709ef2bb0d39267ae927"}
{"id": "train_073573", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "760040f677e44e8a6f86d380bd46bb4f336e4601"}
{"id": "train_073574", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "59c7a2e54522d32c4d2395ae908c8d98a0b0fcaa"}
{"id": "train_073575", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "281ce2e057fa0feb172f3e90adcade717a295466"}
{"id": "train_073576", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "1feda04145ea2a14cd431293f6b8f333c6a89b78"}
{"id": "train_073577", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "ae6c1d940bc2f3dd5e070558e83d265d8147adfa"}
{"id": "train_073578", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "425321b11dfb8d46e2a525d0d2ec15c55b09b513"}
{"id": "train_073579", "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\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "edebe45e98f68dbae8031c13e94521a3ff0adb99"}
{"id": "train_073580", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "0dcdb01c027fb2126fb28f49886127bb22c7bd97"}
{"id": "train_073581", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "381d0c19fcedb55e388daca4de5d3ac2ea1626cc"}
{"id": "train_073582", "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\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "02740d1df984a6decbef5f1adda62039dc6a5467"}
{"id": "train_073583", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "c154ebddbfa1718cf92abcbfe844d0c6984fd202"}
{"id": "train_073584", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "0500a8687c75c74d99b88db7fff6168409162504"}
{"id": "train_073585", "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\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "423f86f82078dabedac61c044f575a138c4b6885"}
{"id": "train_073586", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "4d9899f7dfe82c43ae802740f1e49451cf47e972"}
{"id": "train_073587", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "972325957561ee7e3e61585ee72c436840cc1024"}
{"id": "train_073588", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "5cb2e0c2a60daf0c8bab6ddcb40cbf2ac22522a1"}
{"id": "train_073589", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "ce789b865f393dae3a63c1bb6af1ea938ea3314d"}
{"id": "train_073590", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "bf782c38327a112de7e1b6abea404f90d84283fe"}
{"id": "train_073591", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "d00883c5cc6613444ad1ac83c8e428b07331bf04"}
{"id": "train_073592", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "a6e437c9d709ed20f8efa1a3e91209a0f7e1a88e"}
{"id": "train_073593", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "6502d79da12be17c773583b437f8daf424e3aa55"}
{"id": "train_073594", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "30d7aff1ada2d5986daadd8a7e9466540904e00d"}
{"id": "train_073595", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "c012639704990f535e8727240b5876451b403b38"}
{"id": "train_073596", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "9c1376776be536a1817e8339487a42bbb772ab15"}
{"id": "train_073597", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "25475d8a489813bfbff29de2ba72c1da8121ce92"}
{"id": "train_073598", "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\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "0cba8a8d2707268c64f4e5aa59860d373180f3ed"}
{"id": "train_073599", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "bdbeec50051791b46d10f7caff66b3afe5633e5a"}
{"id": "train_073600", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "8d8bc925cf8204ea011da25b605c0137987a4aac"}
{"id": "train_073601", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "6d05dc7326cd030eb53633fbe80dbfa7ab81363d"}
{"id": "train_073602", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "c9a481ed57525e7b9c9eec428090a7ca056bfc34"}
{"id": "train_073603", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "29990d4f525e2c8c8b06d00ea466223b3fd120dd"}
{"id": "train_073604", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "addd97025bd6f1427c14d912b18e73421e587424"}
{"id": "train_073605", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "ff749d7e95c9b6589821512888315f179cec4f01"}
{"id": "train_073606", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "d877af7fdc864c145ca1dde4f71f98d77ccf7a39"}
{"id": "train_073607", "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\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "8b0ebfdfda7b469ffd42cc43b6600c188cf43df0"}
{"id": "train_073608", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "c1477c001309163f3d168226326f207c6c0dd9c0"}
{"id": "train_073609", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "2ebfa74b82b81e7085beddb90ba1ef04e6a4aad7"}
{"id": "train_073610", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "229b45a2e078c56812f8f360593652cb4294b2f3"}
{"id": "train_073611", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "626cb477fac7e47a44e52258cee5b9d808e78e62"}
{"id": "train_073612", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "9daf6b1b752261710fe494ab075365d7f39d954a"}
{"id": "train_073613", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "bfe568e730b2eb492d222416c94f608133899e97"}
{"id": "train_073614", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "32048c8cfbfc26087f7a1e04de4ce4bfc09a82fd"}
{"id": "train_073615", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "1b8a73c7523f5696eeae40f76e57341c3dd1fcc2"}
{"id": "train_073616", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "288cc0f6acfc86febe079ae4e9865f366d2d1506"}
{"id": "train_073617", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "b3decc8c291975e1c84df6d12270e6db18300a47"}
{"id": "train_073618", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["agent", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "2fe183eebab0db33c8aab932005ec23221e912a6"}
{"id": "train_073619", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "737a0eeb39ee85979ce02fd18a0038a335d0b5d3"}
{"id": "train_073620", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "dc53f4aeffc2ea27fe26a643f2e4d986febd7ff6"}
{"id": "train_073621", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "77e2e402c1d3dfaa5a824daa4cbd15ebadad75ce"}
{"id": "train_073622", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "2372268f71798bc62361b97e6869d6fc6e251789"}
{"id": "train_073623", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "7e9e093ec45e5d0bec84022690956075b77648e6"}
{"id": "train_073624", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["governance", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "0ae50de68556dcbdfde99e7af0ffdbc0f0da4d96"}
{"id": "train_073625", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "1d6b17ffb4a75846ef96780c15556c9ee2ad9a01"}
{"id": "train_073626", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "678acb45823105ce0bb8e85416aef939cc71d488"}
{"id": "train_073627", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "76c52bab886c448be03038eaff9a48e37cf6c5a6"}
{"id": "train_073628", "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\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "203da55c15dc1805f216fed287a6845d8e085bb6"}
{"id": "train_073629", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "4d4a10f1d4204f5713290d64f5b74a90e1e93b92"}
{"id": "train_073630", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "5d3f3e9073fb7e9178ffcb3f3b6f44c5b1b4a5ac"}
{"id": "train_073631", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "fecb54b2690a1c1d1c34c27e90c0f91d87339cc9"}
{"id": "train_073632", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "427e7c41854dddc4077c2f3770d1fa2a2707cf5b"}
{"id": "train_073633", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "1c391b43a944d8fed824e252e25d23a74cac7ae0"}
{"id": "train_073634", "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\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "6aea150fe17a1b6b3ab159bad4f31692d71e1b81"}
{"id": "train_073635", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f78cafcdbdb1f7e3db28f682d48f7c1084e4d4e3"}
{"id": "train_073636", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "35be5061a305be125e8772fd636bceda7d62d76e"}
{"id": "train_073637", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "1b858b4a13c1b915b47be7b1e46be4b82f627c45"}
{"id": "train_073638", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "eac7428f5cfe2ed21e6a00201d93850c355579c3"}
{"id": "train_073639", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "6ddc33e78b4bf43d07f4c1fb5c9a55efe9089245"}
{"id": "train_073640", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "edc5d03752ac7c0fde8e2bc5f484d40032aa4c0d"}
{"id": "train_073641", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "2957d0db4bfa5861c5560af8bd341a507776f002"}
{"id": "train_073642", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "b93c2eaa45a7597e49eed196107305d5cf9f7500"}
{"id": "train_073643", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "733d53baba355d96187746fc4901949dce111cdf"}
{"id": "train_073644", "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\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "1cde996d8a4ac5f366b6978db318cecc440e0a2b"}
{"id": "train_073645", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "13037692420fe31838ea11cbe3ca0f2db7320807"}
{"id": "train_073646", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "83a491232d63fd369bcf000931d4ab515fdad54e"}
{"id": "train_073647", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["research", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b7e9f03c3e158b89584864f55c772b5598252952"}
{"id": "train_073648", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "a24fbfc64aef613f9588f4bf8d1e926a2232c94f"}
{"id": "train_073649", "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\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "1db8a8834db874ad911cfefb1914f0cb1d533b2b"}
{"id": "train_073650", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "62949c7c15652a54d636060dab82da713458fc84"}
{"id": "train_073651", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "25554a303befebfd4fe0620d0751a856bf558353"}
{"id": "train_073652", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "7b4f4a94580028d2e2120365e449dbe66178a045"}
{"id": "train_073653", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "a3f7950a8dc33b5103121d1fc5f8cdcd2a555c68"}
{"id": "train_073654", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "88d226319056df084c792e28ca1cc47c00817066"}
{"id": "train_073655", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "b36570fc00006be0505adfdfe1638f919efb6c42"}
{"id": "train_073656", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "44a8afe560771e710e3c5320f279bc1909af61a5"}
{"id": "train_073657", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "00ce97f88bf8f9b014cbfc120996d0d06a5f78e0"}
{"id": "train_073658", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "13eb5e9c07b6b20896a37ad65067d0b6ade40579"}
{"id": "train_073659", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "b2bcdd4fc0ab51b5b3908dec874e39497651fc09"}
{"id": "train_073660", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "d397776fd1d08242c384085570c420eb3ab13b66"}
{"id": "train_073661", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "8374e7ff2d63e156023fb7e47b282ef9da0a5e94"}
{"id": "train_073662", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "3a8881cd02274169dff9dd4f75ab246991995b2f"}
{"id": "train_073663", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "1d4061b0afb028f3b9b12f45113691dbfca92765"}
{"id": "train_073664", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "1a59ca3cbef91ef2fb8a3f49494e79ac5b24b06d"}
{"id": "train_073665", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "d783ab821be2019d5a7449dbfeb35245b4adaa04"}
{"id": "train_073666", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "84c407ed32541f8d89218a4adc04d88cc5a3ac32"}
{"id": "train_073667", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "3fda00a5158fcf815ee9fe245ab178c1ff0acebb"}
{"id": "train_073668", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "e680755142a6f94d5ebc39a9ce5b9484a5b4da20"}
{"id": "train_073669", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "b969b5cb3b9df5b3c27d654748930598e48bf1d7"}
{"id": "train_073670", "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\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "e6a90ff22ecffa4a0e30e31714ec2b2317e4308e"}
{"id": "train_073671", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "a28f5bbdec447fcab623a2349ebe63a524f0551c"}
{"id": "train_073672", "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\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "00ca8b239c567cba5a4a84e9d43e894450b5ecf5"}
{"id": "train_073673", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "31831699f3dcd9c121b4be6d73cae221ecbf3e6d"}
{"id": "train_073674", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "2117c4da758236600aae5bd37542b353887601bf"}
{"id": "train_073675", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f3270b8684c8f9cd39462e3eaf3b96dd09c2cb6a"}
{"id": "train_073676", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f7ae9d90912c8d1e7ae37b4d3af5ac2031e70e58"}
{"id": "train_073677", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "b3966054db6265a57dea9c652ff274df98d96ec9"}
{"id": "train_073678", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "4b2bdd038ad40d7ce921b49dbb9e7385a757cfe1"}
{"id": "train_073679", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "b60b0b96cd5839e8592ef5870cac7d13dea8c333"}
{"id": "train_073680", "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\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "91276eeba7da5751a32aee0172a48a5d391bb24c"}
{"id": "train_073681", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "3d47c115b9adb95bab05647b04abd22790b896a6"}
{"id": "train_073682", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "bf38d9e9c7291e8e1887f219ab1fe953012e9c58"}
{"id": "train_073683", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "59245ce78be830f009995abb768a5232df7fd53a"}
{"id": "train_073684", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "7ab0a2de81fdd6363f62655faea67fbfe03583d1"}
{"id": "train_073685", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["agent", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "12195d722b6df3771c523c15c215bc7f7b920a87"}
{"id": "train_073686", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "5e1cf53aaaa1fb61b68962db18f7a588832a8a8b"}
{"id": "train_073687", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "6d144fe0e68ac7183a17862d95ce12814d6e28b1"}
{"id": "train_073688", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "6dffafee87439105cc89c5a5561e2597eb7b1ad9"}
{"id": "train_073689", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "22eba55367b71252bffa1bb6b5706181353bb854"}
{"id": "train_073690", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "ea5566ab3fb701b105695f6c6cdbca4e9cd0a001"}
{"id": "train_073691", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "db361ef9a69f7ffbc0b43319dd1ba297bfa7471c"}
{"id": "train_073692", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "5ff8a55f5217e205ba33e2d57f77306407c8d0b4"}
{"id": "train_073693", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "8404805c3d317276d6344fea20427cc34bebebd6"}
{"id": "train_073694", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "5795d7df9ac6c8fb5b4bf8e51d86fef7c90eb5c7"}
{"id": "train_073695", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "032756633a9ff6108bd44e3f1daee9ff2448b5ff"}
{"id": "train_073696", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "0ef3a293381b46d079005f84c1eec110908088a4"}
{"id": "train_073697", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "41ed2c297aab97860cde6ca6ead9850fead4db42"}
{"id": "train_073698", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "e37352725fa2a61d2b16d233ca164067b1602535"}
{"id": "train_073699", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "58c22a01ebaec8e2f5a43fdc3673ecf47e739d67"}
{"id": "train_073700", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "581a0fdc348e4948040c0b994dfb21da92b17209"}
{"id": "train_073701", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "496bf26a95da44321dfbaa93fb06cc06f15c4aa7"}
{"id": "train_073702", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4f128ace204a8041c6d993b44c42bbe39ec68c97"}
{"id": "train_073703", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "f4ad8e7b329e44edf63be41f19fcc96c99355e70"}
{"id": "train_073704", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "98f0c001559abb28167e8f55d917d0d8efb36bf4"}
{"id": "train_073705", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "6b0c8d0c55b2d9a2ae6c9409052ee243a3c92799"}
{"id": "train_073706", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "04eb598cdeba1986a4127e9c6961de4761ea6929"}
{"id": "train_073707", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "6abe308d8dbf287a7fd5f610612e3959887e0952"}
{"id": "train_073708", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "a42cb5e0ece5e60b073c4a8b287260d894fc3a36"}
{"id": "train_073709", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "ab1f01775266639a5cbc55c72a78efd97479eeb3"}
{"id": "train_073710", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "0ec0e424d2b2728e508e76c9c5f12720c40cd150"}
{"id": "train_073711", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "8e508e8439728b19c1f8e153709ade1f305e585c"}
{"id": "train_073712", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "1e3e57c272156fb163e6f28e8bae65b4117a96ce"}
{"id": "train_073713", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "ca46a3ae1a620460e94d42e8015eede3e93e26ee"}
{"id": "train_073714", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "f30d16c4c6b23bec248b600c4057e5afa74383b4"}
{"id": "train_073715", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "41791728e1db3b6ab26e58d39eff9c9b41d8a0c9"}
{"id": "train_073716", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["agent", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "6ab6948c109e67e9d6bfea0337141db33beb7861"}
{"id": "train_073717", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "7d72fa70eb3f949d0eec909695c96a9676b940c1"}
{"id": "train_073718", "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\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "5062806ef2ec17cf7147d902f87dc5eee18bebcf"}
{"id": "train_073719", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "50e739a96d6b4cf6b762d3a6fdc59ec838fe8fcd"}
{"id": "train_073720", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "c46cd080e7bb5bc80c98546c3f1372e577198dd3"}
{"id": "train_073721", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "c1cad9ac1d09a4f065ae4c02cf48e3e03c8cfcc5"}
{"id": "train_073722", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "ed125307f0f99b71c5479f8f290fd5804f3f0b60"}
{"id": "train_073723", "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\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "33c238c4be49a3e301c5a7d8e2fb282c714b696f"}
{"id": "train_073724", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "8b2810cea2c34f5954a2811e127eca20b84c539e"}
{"id": "train_073725", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "3124eb984ba349ad49719d6beea7f4874befe4d1"}
{"id": "train_073726", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "60420fbcd8b9d3982401faaf6dd5af0ddd8ca6bd"}
{"id": "train_073727", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "7e42650e4f8b26c6117b0e45e96d90e1c4d37078"}
{"id": "train_073728", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "e9d29b7bc90f2d4b722d05295e1f6ad55e459b02"}
{"id": "train_073729", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "5a8f8f4ed66ee4b43f08834e541c8b843ef922bc"}
{"id": "train_073730", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["evaluation", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "fa59a0baddaa6b1cee306fdf2a76ca392fe4b835"}
{"id": "train_073731", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "5debd765f9de29ccf6e5fb0e84f19930a525fa6a"}
{"id": "train_073732", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["governance", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "7ecc9a0426d1c29e06b49d32ec5981cbe7756a3d"}
{"id": "train_073733", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "27dfd2ab45d5f0ad233f564a675853eb33f080fc"}
{"id": "train_073734", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "4c2d7e98302e1a276b7ab725b2627481d2b6f1be"}
{"id": "train_073735", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "10cf5f8fa3e3645893f7bf88cbdce8944c909a9a"}
{"id": "train_073736", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "447755180261688e39a205f46dbd789b81d54e13"}
{"id": "train_073737", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4efc958436354093bb6fb6f3e970960ebc9f5f11"}
{"id": "train_073738", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "bd52cca696f51c96a25df6a13c2f4a79672a9f9c"}
{"id": "train_073739", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "8b1c0b1622d24ca7d65cc66fb66350a555af24de"}
{"id": "train_073740", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "c3abcea63e4353d0e0811838ec2309c634055224"}
{"id": "train_073741", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "c2313b35e965fee0070c06cb3ede4c89639c95a7"}
{"id": "train_073742", "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\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "378126deef11d8b63402d63a695ef78b27f1573f"}
{"id": "train_073743", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "a086921b13d892bfb5fd55885c7da5f6509278da"}
{"id": "train_073744", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "4ee74b3f12a3d3536d0e2cd9943c9d85e57e4bed"}
{"id": "train_073745", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "290da7e2caadd9c61617c64348a3d0b280715e32"}
{"id": "train_073746", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "fd28eb56f0daea90015bca7fa23aa7c81b845a43"}
{"id": "train_073747", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "372b5008b0937b6a7becd6629c111cf877985e2f"}
{"id": "train_073748", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "de01425e5288a52a1f298c434372cd3faca25692"}
{"id": "train_073749", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "be479522072426273463829720e9f33b6e126230"}
{"id": "train_073750", "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\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "e3d3669fb119330b31bf1b78b13a3b0050fb965c"}
{"id": "train_073751", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "89934cd3df295002cbcfa81ce2df5b29f13d13de"}
{"id": "train_073752", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "39ed88344489a4e822ac85da450f46bac25e44f2"}
{"id": "train_073753", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["security", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "80aec6b05e14c68904e3159f775e800319377166"}
{"id": "train_073754", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "8638dfce1e9dde4c4701b122a09c32714be5dc5b"}
{"id": "train_073755", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a8306381bf0cda411f7cd30171c9a0bb39d80a92"}
{"id": "train_073756", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "bc961c94d0bd298a0ec0b529df5a37d6396e816f"}
{"id": "train_073757", "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\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "27e32673cf36ccc5b627afd77d5de16939095ed6"}
{"id": "train_073758", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e9dce1ebbab75b51a441d58eb4c59eabf0cdac0c"}
{"id": "train_073759", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "f13377851072dc02ff49645f688856a927ec661c"}
{"id": "train_073760", "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\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "791e7f72f39fddee7bd68a28223ecb15cf25cf79"}
{"id": "train_073761", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "96d10e67ebda9a25ea51dad6613e16a080c29135"}
{"id": "train_073762", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "27a7de87a08122a1f4af1c5fce8953b2f7d764d1"}
{"id": "train_073763", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "8f7a3628a73ea1b612eedac6265b80a87345290f"}
{"id": "train_073764", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "8f4904588fa27d8ab54b9ee15ec00549c5f59da0"}
{"id": "train_073765", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "4a6e7b0996619281467fafb27e6e0d71428145f7"}
{"id": "train_073766", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "50b74975d083a71ee7804c9defdf1a75b02957f5"}
{"id": "train_073767", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["agent", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "3bfc5198b39df97e5b7ba549dd63f5f6e561647d"}
{"id": "train_073768", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "4a7aebb47b959412f19429b5c302c688bd6069ec"}
{"id": "train_073769", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "2ef83af3321b8ce65b2625490b3669f1bfcc4386"}
{"id": "train_073770", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "ef8f8b0713a1a5fe0e739d83abb186d362f66cd6"}
{"id": "train_073771", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "8baa0bb7546d01bfb36706e14f79551294f63680"}
{"id": "train_073772", "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\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "af0536687d659b46bfe66aafeaa9644d974de42a"}
{"id": "train_073773", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "b86c5fe7d14bf6da58fff644c6d7586b51aeb22d"}
{"id": "train_073774", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "3d364663fa34b7603dee80e7b3c58c6ea21c1b57"}
{"id": "train_073775", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f18fad4f73ccbcda4b2764cf44a144a829595a8f"}
{"id": "train_073776", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "6a54dbc4aff1d8985044fdd45dbe87daf7265f41"}
{"id": "train_073777", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "99efa2484c2f32da3976f52f97d64f5a9535b7b0"}
{"id": "train_073778", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "cd038faf2135e99eb3a0ec40a68ecfebdb8cd28a"}
{"id": "train_073779", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "7f0be8ef84b33aba7dff3ded5b7b0624234b98f3"}
{"id": "train_073780", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "7598e72094527140025e7d75d4faf2998e3f2af3"}
{"id": "train_073781", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "40bb2c8483fac63641110f70cc1a452006049819"}
{"id": "train_073782", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "fffd6cddab5a191718495f789913c6f88c082a2a"}
{"id": "train_073783", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "e12a77bb0d33bbba92889d5215f5fbc8290300ef"}
{"id": "train_073784", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "7c90998d0030a3e0e4ba65f6c0706a342a8b45e6"}
{"id": "train_073785", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "2fccb0c5deab1b293e3a067f96e32a212dff9f31"}
{"id": "train_073786", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "260dfa8091e0010501051127a204219b60b2a091"}
{"id": "train_073787", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "888ac850bab3ea07b7f5e3fc461ea0eaeb2469a0"}
{"id": "train_073788", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "88d226319056df084c792e28ca1cc47c00817066"}
{"id": "train_073789", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "2ce4acd63e113825f2f80f4646f57e3b72669e5a"}
{"id": "train_073790", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "27f5869b05117b8657bd7591fbdcd858166f7b18"}
{"id": "train_073791", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "989627840ac593f619cee67af6f5b218dfa039af"}
{"id": "train_073792", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "23e91dd97ca412201230ece69adf651c83c29682"}
{"id": "train_073793", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "d31c8265d72550686fa655e561e3864ad2481ad4"}
{"id": "train_073794", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "83ce991d5dd8baf597741a5d1ed22089433bb271"}
{"id": "train_073795", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "4293c3d0aea19fcec2f65d68a193c158a32c8890"}
{"id": "train_073796", "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\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "46d4eb52fb0755c85957282eb242ed77456259e6"}
{"id": "train_073797", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "5dd267fe5e55d27fbd3eba0d6a94d487226f142c"}
{"id": "train_073798", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "5aad104eabd7aff882d76ea4dfe1a7e9ef0ad04e"}
{"id": "train_073799", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "1826b134c1eb4a551f47468fed6a4fed59431abe"}
{"id": "train_073800", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "42639df93cbb6ccc762fdb70399650d4002d5231"}
{"id": "train_073801", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "20a4cca2daa874407ecebf065d3380f07406fbbc"}
{"id": "train_073802", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "c1465ca9dccd88798d53f72e45380ca3b0ecc673"}
{"id": "train_073803", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "69ce7bda7e1f1afca9b8010192723205e0397cbc"}
{"id": "train_073804", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "6d9c331be43c3274909b6fa9ebce144ba56f88d1"}
{"id": "train_073805", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "aa06d2e4fa8a3658cbc9bcde2ac1eb20881d50fa"}
{"id": "train_073806", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["governance", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "136750d260c38d88348afff5f9be67a0e2b4abcc"}
{"id": "train_073807", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "dadb2ce42673cf789d4ab911727e64e414a43eb7"}
{"id": "train_073808", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "6885053916035f23b2dff4cbf9341d260a9b986e"}
{"id": "train_073809", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "ae26be2023d8fd4f7013171a764dcf14d0321cdb"}
{"id": "train_073810", "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\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "85c603223803a624facf701523adac5555a58f7c"}
{"id": "train_073811", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "72b5d0d6cf5ab900153c1c3e9e38cbcfd74c3769"}
{"id": "train_073812", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "6b393496ca8fec8334ba2626e0b114c5f1741adf"}
{"id": "train_073813", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "dc68260f97746fc7b4697be1b5442ec10be85176"}
{"id": "train_073814", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "efa4e123044ca400bb4fd3179f4e539bbf142dc2"}
{"id": "train_073815", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "262c27b6ae0a18bd3e5400bbac494a6b9fb16567"}
{"id": "train_073816", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "289a7568d408fece2b28cee89fc7407a2155eda8"}
{"id": "train_073817", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "ed3c0e0e5219424ae82d92756d317fe035e3fcc3"}
{"id": "train_073818", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["research", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "82a65f2d5f9dc7743f97505c6a72a707eae90f62"}
{"id": "train_073819", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "d08dcf38e095f9d2e9fd2688f46ce9a247151755"}
{"id": "train_073820", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "3f3263ed2e7781ee1f22c3423944c956c8033dd0"}
{"id": "train_073821", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a5bb6fce36b558e31ea139f3bdb408d37411cc48"}
{"id": "train_073822", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "aa901bc34913c7f68582475d2d4a39308f09e8a0"}
{"id": "train_073823", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["coding", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "55d876613fede9d4611d53ea20e23d38ab40bbcd"}
{"id": "train_073824", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "4408a8ce6f55f8ac1fed3b171d52edfeddc0aee7"}
{"id": "train_073825", "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\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "52738b62546ebe96f4b15b88125040a9aec8fa96"}
{"id": "train_073826", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "2bff81d98807c7a456918b33d732c52744c86471"}
{"id": "train_073827", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "dc7980699929340416d490334df26a8227fab3d1"}
{"id": "train_073828", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "705daa25c9528b31a93e8d3750d10490d16ac1f3"}
{"id": "train_073829", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f53972549c1445e30ffe336dfeccd4c2dad39a9a"}
{"id": "train_073830", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "c434dd13a6bf6f71a4e420bf8a665918891a0622"}
{"id": "train_073831", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "315f168898838d7eac203876a9e8df2e5e9fdd4f"}
{"id": "train_073832", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "2f74b9ce9dacee3eb1b9937a2e64b5021a01448d"}
{"id": "train_073833", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "412e883cf7df0cfe5d5491fcf90827868c189672"}
{"id": "train_073834", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "8df13624910f7d13ec8c018f80c7263fe07c0aec"}
{"id": "train_073835", "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\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "187ef2462c6b6cda1d739c0f14fd2c9156f78dd1"}
{"id": "train_073836", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "380293d614ab443e2dd518c4eb16a4adf66783b3"}
{"id": "train_073837", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "3d408bb18f315ce30725f226f3103e3e7addd271"}
{"id": "train_073838", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "b0816926d3a12abb764f63e4134e41ea32f170d3"}
{"id": "train_073839", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "c358d731ac518f01cba27e5dc2cc8e392ab94ef6"}
{"id": "train_073840", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "bf1d0e051d53544bc9dfca69dc68e1532fcc8a7c"}
{"id": "train_073841", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "2c8bca35f6e2e5113e59040e3045432f921cc63b"}
{"id": "train_073842", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "4d19544c57d92d7cae2989f1b2aaa3a1a286b62f"}
{"id": "train_073843", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "f76bc4a772d07d21b5e821457936f0b12a91df29"}
{"id": "train_073844", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "086bb3e442e3cd69a9edf83e96073b47c79e6b64"}
{"id": "train_073845", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f9ff0813cca58c052b1236640ab52d7feebf0a97"}
{"id": "train_073846", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "1d517676754365328e1ea5c352e3ad912c404d10"}
{"id": "train_073847", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "953eff5cf85ae3922fb9864a6f3332df1d2c3e07"}
{"id": "train_073848", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "ab4172cfa0f3b7e4030d0822d8aba74e46e0bd76"}
{"id": "train_073849", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "4bb097aa2bd9c6008beb364680f9553b84426054"}
{"id": "train_073850", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "443b5b1f9abdae35552c215ac78a892090f00be8"}
{"id": "train_073851", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "8a424b249c4f5b010541a7be37adc80ea3668f53"}
{"id": "train_073852", "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\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "818c047248c85c96b58d8290a557fc02e035b916"}
{"id": "train_073853", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f0bcb6c15fc901eea6b4493b53c67c3ae6ab9421"}
{"id": "train_073854", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "af63c90a6297154c04e1c426669783e729531a85"}
{"id": "train_073855", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "30ef96b77b8079c5ddf6f3055621c0cb513c70ec"}
{"id": "train_073856", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "871d5ecb73722b715f279be2905cfd9f448ef4a3"}
{"id": "train_073857", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "0c637d7c31f007706dcccf84bc2855a9afe3e18b"}
{"id": "train_073858", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "9052e8cbee24c1a72be027fa96a57c3c7bf80049"}
{"id": "train_073859", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "11c78b041c6bf4c6f893fcdd23b6bcb5b8093482"}
{"id": "train_073860", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "6182d67a948536cf4c9509ced0c4a72fa699f258"}
{"id": "train_073861", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "404ea1aadb6c5791f4d69cdb13e66b2a87007106"}
{"id": "train_073862", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "77a4c791b629cb23ee06a6227b23a1bb533bcd7f"}
{"id": "train_073863", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "2d92c400bf3dbc0e15f404642de874807aae732a"}
{"id": "train_073864", "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\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "6bbb2d9f20f2d14660b117c90b520472a147c9b7"}
{"id": "train_073865", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "994e41a627ce2c8206a56cceb097d1b7412e9b6e"}
{"id": "train_073866", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "dab9388e774b9576cc2460402a4115c37c1ff3c4"}
{"id": "train_073867", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "d909f294e63f77342ae778ecc20a1d3a20cc606c"}
{"id": "train_073868", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "8c918da15c81d08671f8908741e6ade6881c0a17"}
{"id": "train_073869", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "ce8f31e33dd3771f2bd9a9830e818894bbe21845"}
{"id": "train_073870", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4cad9033df2ac308f0a847b07e6716e3e44f744f"}
{"id": "train_073871", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "4bf4feed70744ca0670757a48629a4c56a7cf7c3"}
{"id": "train_073872", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "fc25ac099ca089f646109e2a094a067dd3b89e84"}
{"id": "train_073873", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "ae09d459cae1c380cd59a27a8bc74c67d51c733d"}
{"id": "train_073874", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "7acc8fa431e884ae3866d0734198dcf21eb22f63"}
{"id": "train_073875", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "38d0c53c3a77e66328675f021a4391ab8020bca1"}
{"id": "train_073876", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "24cd7ad76a0621b6a5f91e8446e12267593d04db"}
{"id": "train_073877", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f82df0fb7782e1ffc914be93d5596b2329f25a0f"}
{"id": "train_073878", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "926b7256cd3ddbf91addf3859f1d929f6470e0a2"}
{"id": "train_073879", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "b77b5fda82532730fcdd3732f0087351e0f8136d"}
{"id": "train_073880", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "ae32b9bae03142547db515d17f6d68aa63cffbeb"}
{"id": "train_073881", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "0abe8ed8ae83f44194a7c06b808049bf280e5188"}
{"id": "train_073882", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["security", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "9398efad2508dec954078fc940287838c59eadc3"}
{"id": "train_073883", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "1abe2b2c056152d8d4ba6a23c651e76b383593b9"}
{"id": "train_073884", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "fd5398acb7b945e0970491aa3a698b49a9b0e78b"}
{"id": "train_073885", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "4681bb93f966b7c7624c84bdb08b662cbb0564bf"}
{"id": "train_073886", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "adc71ecb3d1c724fc2c75c9504dbea357bc64d20"}
{"id": "train_073887", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "7d62ff82b1129e3182e45dd79edd2f574a3446c0"}
{"id": "train_073888", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "0e41438e2fdc8cce8dff5bad1f66a0c7b8aa6056"}
{"id": "train_073889", "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\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "0d0eb90d2d18878c2f9d1155cc4566ffde663d37"}
{"id": "train_073890", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "83c6dfec8698bc2f820925355ef1ceb3ae702140"}
{"id": "train_073891", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "fe1e5796e5780bd89903ee30c10723c302cfb2c5"}
{"id": "train_073892", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "812734a69055a3e0984e219a2b791f397c156ade"}
{"id": "train_073893", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "9f76a7db78be1cfbaf0c4e3cc7e5aedce0cb6507"}
{"id": "train_073894", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "62e86d2da29f02c81bace762f26ae8e5bc73812d"}
{"id": "train_073895", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "35d9baa629c894f66c7965c5efca936a75317d43"}
{"id": "train_073896", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "5370328fe2e580dde8740fe250361a0158f976ca"}
{"id": "train_073897", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "d529e431526629c0b5d00578db55aab9a8e696ff"}
{"id": "train_073898", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "7434996713d7ecc7f491b498bbbbda5e0a5a4722"}
{"id": "train_073899", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "5be03b0384ca5caa3cb95762d9ea4f4f4080126b"}
{"id": "train_073900", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "69a7735bba98017c670d75c5f9ee9d7be1de6a21"}
{"id": "train_073901", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "9916cbd38ab1ebe6e30e69f43370c750ede438a7"}
{"id": "train_073902", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "2ec6001b62c95241772c67747bd9f4b72e791bfb"}
{"id": "train_073903", "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\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "1569936c8970112ce1e299ed74e09ca027fa38f3"}
{"id": "train_073904", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "37f494b722ecbb3ba5f44ef66043ae72ebc6b59a"}
{"id": "train_073905", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["coding", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "19357d617c73dc462c39a21f54aadec93ab37b98"}
{"id": "train_073906", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "9a4804d491b6537c1daf1da23a23d0926acff5fc"}
{"id": "train_073907", "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\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "2ab17f50df5ee830d501d849dd9d3f6f43c3a24d"}
{"id": "train_073908", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "7a7782c7c07d61d600e7b082cf93547a621090dd"}
{"id": "train_073909", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "a615343870b1d7f7676c7b1b050f6372a4f9c60f"}
{"id": "train_073910", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "85c490cf7c3f0e5c326bec8f56b618fe526a6966"}
{"id": "train_073911", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "1510a925ae4cd6f87ffdc0bf172fc812db724c89"}
{"id": "train_073912", "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\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "db2bfc36a31f333a50dbed2d990f76cb0c14af6d"}
{"id": "train_073913", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "855655c2ea785a6bc01668dbccc51da552fe5bbd"}
{"id": "train_073914", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "32d3db6e817a3adcf70d5b88f49a7da8d54c2a3e"}
{"id": "train_073915", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "cf29b58df72fa4e756ec874f596df644651e8590"}
{"id": "train_073916", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "b8b910aca9a7081194588b08c81869f4b092cf02"}
{"id": "train_073917", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "9207cd35bbdcf1c0d1ce73fa912e248d14834a64"}
{"id": "train_073918", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "13b10e706d7cf5a7c6b9ab3956abc28ec34f5e7c"}
{"id": "train_073919", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "85558d4201e13e92ddd44089053da4cdf91d91dd"}
{"id": "train_073920", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "e08701f543b99c21e7db5c25fdd27b37ecee28ff"}
{"id": "train_073921", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "051c297a852cc060a270c29865d666261d790e39"}
{"id": "train_073922", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "b9c269d1397fd27fb259c7414b6cbf1a1bca4bcf"}
{"id": "train_073923", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "0d712c993ccd69fd493c7e3450dc18e1027f9071"}
{"id": "train_073924", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f36105498e31936f17a1f3b6ba47543c42bce675"}
{"id": "train_073925", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "0e388492996235e14413974aa222775fd9134092"}
{"id": "train_073926", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "273c75c945be9cae72cf17796b40524f4d8ac638"}
{"id": "train_073927", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "7d036c28559e6091b60edff3a61c2156014572c9"}
{"id": "train_073928", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "bdf0414ec167a3afff51a645c6c8b5806a37cd6f"}
{"id": "train_073929", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "8071a8086c7ac04ac27a595d01d1d0b00402f035"}
{"id": "train_073930", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "f7305f02c60485fd549d1be8c0660c451aa1b414"}
{"id": "train_073931", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "84b862e3d9e9d0b53641e3f4d3138ab9f366f245"}
{"id": "train_073932", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "46419214383bb63e9e575f72f2846327f9b3c5a1"}
{"id": "train_073933", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "a9c2d2678e9770ab38151b9e54a4d29db759826a"}
{"id": "train_073934", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "3ec9c5b454ab970233ffc0fb64ddb5d3f20f8e04"}
{"id": "train_073935", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "76c9892c2bd8e61f0838cae655deb0d9803afe1f"}
{"id": "train_073936", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "03f1c1c83ef93c4e2bf3bf1e0207e388513e2edc"}
{"id": "train_073937", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f68a38e5220fdd6e3b52b6c98a97645546e300a5"}
{"id": "train_073938", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "84bd4388c2f92075eb488c704e0b8a2008d61463"}
{"id": "train_073939", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "82b03c12f982747cbc1e2f74ead3fe0b9ffc4e45"}
{"id": "train_073940", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "4f4befc7e55932b5bd0b6469f6a05d960f297bfd"}
{"id": "train_073941", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "0d3dfcebf8124c413130568f206124ab097d8915"}
{"id": "train_073942", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "4255382c306fde439a6ff9b30ec5ffcb2d91648e"}
{"id": "train_073943", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "67641524e5f43678e8d32a3a13802f5969a22bdb"}
{"id": "train_073944", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "de1f169d02d7eadec03a310643bafefd2f031a07"}
{"id": "train_073945", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "898167b565a150cbd559e48052bce7e5f8b1647b"}
{"id": "train_073946", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c15d22d2eb0a4bbe877f0be52db6e4b5c944f6f3"}
{"id": "train_073947", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "ee271beb1352fe16c99c59411a23535e03101b4f"}
{"id": "train_073948", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "5f084392c794deefb7e069a3219f7e777f31d954"}
{"id": "train_073949", "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\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "75f205a378bfdf3c2d2177e4fd56fd74476c0e46"}
{"id": "train_073950", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "91e738f7a37277455dfb587d3358d8f4800f877c"}
{"id": "train_073951", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "c264fa2091438a81262eedd27aea7fcc49b70055"}
{"id": "train_073952", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "0d8521f8f3dd38f7349b49e22619c67a6a1ebaee"}
{"id": "train_073953", "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\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "31394e369d6351d18d32c563279f8721a77a9e78"}
{"id": "train_073954", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "5f39b46d7e36670bfd9aae875a793ee3eb6dbb17"}
{"id": "train_073955", "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\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "14be10890e09b5e81cf79c99447878a1a0b9631b"}
{"id": "train_073956", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "ea716d28335c4861bbcce4ec36b8f4a9ed99fe23"}
{"id": "train_073957", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e902ba1b0f2eb66d8c1704baedecc84956812a91"}
{"id": "train_073958", "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\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "c93b2b3db404018f8fc5fe2fd33173d7ab608933"}
{"id": "train_073959", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "d56aff2cb687377ef9a2e9cfe64677f27ac26ea0"}
{"id": "train_073960", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "03d4ef4b9835405bcc8d58d2a84d21d905fda6d1"}
{"id": "train_073961", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "b35cd65f325742a3744a181c2a22b594e714477c"}
{"id": "train_073962", "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\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f784f2960db226a72b7c5e63c0b9fc84d544d53f"}
{"id": "train_073963", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "da9deb67d7c8176a6759b66a87a23a806d9f298b"}
{"id": "train_073964", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "20e09f2d4b0c4664043dad12951500eaeec8198c"}
{"id": "train_073965", "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\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "484372c47e581e062761335b926b7cefc5515ee3"}
{"id": "train_073966", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "38673fd95e5a049ff22941b23baa56732a9517e6"}
{"id": "train_073967", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "5e03356b68cbb94eb2c5c9d6bb87ea4fe081e802"}
{"id": "train_073968", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "6320ed841e49acb45d7c99bbdb031be2925d7520"}
{"id": "train_073969", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "625a381c8e6dda56526688461af7f6af739b056c"}
{"id": "train_073970", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "eab556d42943bed383a13b09781870200bfd5301"}
{"id": "train_073971", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "7f29cac5703b8e51f56f140e458a7da405208472"}
{"id": "train_073972", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "bd6fb5b87806e7ed8eb656db6566c6c0f41ee4bf"}
{"id": "train_073973", "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\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "1b9b571c605f5f6f7f4824ba56c05446254481ec"}
{"id": "train_073974", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "9edffdab2dfe17bdea443c6b673ad105cb23ce6f"}
{"id": "train_073975", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "8961118b61d6781ccde48162790ae04671879ee1"}
{"id": "train_073976", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "2584524c73b69f93eff21e6f2a48ed29849a79de"}
{"id": "train_073977", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "190724bd9b7bfa6f64a2c6585e10bf493eb334ab"}
{"id": "train_073978", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "00660d5b4761c09a2f8c2f0ff898a580a2ed1181"}
{"id": "train_073979", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "1c60ab7a201df4679a6ed7244d3e65e6f388fec6"}
{"id": "train_073980", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "32de34ae3e4322e6782d2ad8a215b95c1c967e4c"}
{"id": "train_073981", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "f7de22fd3c087950e422b0e652606c9b94f8baa8"}
{"id": "train_073982", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["governance", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "71207c1325f9576beb398d377bc2bfc3afe1c22a"}
{"id": "train_073983", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "6d5933467ad2b4786a498003b03aacd9ac7105ab"}
{"id": "train_073984", "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\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "1bbd7fe3903a7a55055449e6d09d88701e749c44"}
{"id": "train_073985", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "79da59fe52d33b00ebcb77f67c88a7a5afa52fbe"}
{"id": "train_073986", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "79200cc278aa548c8758a3f8cd7bb9ac9e227ddc"}
{"id": "train_073987", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "7df9d0678aa76ad5779845388d1d2607db1cd48b"}
{"id": "train_073988", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b4f98b0c54a6b3935d3ddadd05666a4a42fb6ac2"}
{"id": "train_073989", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "f7584da0095f0392128a77162ba12b898b9ed351"}
{"id": "train_073990", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "35db3d223e108e90fd6f8380e6a2a19319e96c4a"}
{"id": "train_073991", "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\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "9f9d9e0248b63441ef5af60f7292ec3842fbb0ed"}
{"id": "train_073992", "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\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "14228746434456c3e54883cad34d17e834c54afc"}
{"id": "train_073993", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "d52d73df8076eea629e68901cb8bc87c4fc99b21"}
{"id": "train_073994", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "5093fe04ae6ed227f7dcf371cb12b222e3346444"}
{"id": "train_073995", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "c214a79f86152ed1c00941dd6af624ecd54234f6"}
{"id": "train_073996", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "843dc72b06d4c5603aa4f28c779181b1d2a60b6e"}
{"id": "train_073997", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "38dac40732734c769e0e715fa457f43ce70dc1bd"}
{"id": "train_073998", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "4b3274a4f36d5f11bb2371ba548bb4fef25f0828"}
{"id": "train_073999", "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\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "1b94a7850cce5419e1143edf8d609cdeff1a5d1d"}
{"id": "train_074000", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["coding", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "b7fe3971fb90d7475894bbb6ae1e86c8781b2dc9"}
{"id": "train_074001", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "bdd72856597bdab9f988d7172e209fbfbc6da6ab"}
{"id": "train_074002", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "748a6e819da525a75a09e24322ae425f1c8703cf"}
{"id": "train_074003", "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\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "a24b3f285cda550dfaec313f86181c003147e3a3"}
{"id": "train_074004", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "0fb308893de6eff21775960b3173a9e59f5429b2"}
{"id": "train_074005", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "7c1161c7f59118faa01a5e5957fbe2bc88939640"}
{"id": "train_074006", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "1c7b3f2a33c81c8615685c09e656020a52ed439f"}
{"id": "train_074007", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "ce3740cbbd60c69a99294c971b2d772fda2bcf23"}
{"id": "train_074008", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "d5dd23df9d2ae5ba4b8aeb94987a17e9ff735e47"}
{"id": "train_074009", "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\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f0babb971f52704f9392cb4455becf808eddda81"}
{"id": "train_074010", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "37d3cd971077247bcd428062ae1ffe53b8d8992d"}
{"id": "train_074011", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "6c9ca3dcdbe34dd8a2cf18920bc823740f0df014"}
{"id": "train_074012", "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\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "dac02a1a978c39f81e1cfe2b5b820f66352c3de0"}
{"id": "train_074013", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f6dbe0050dd402097820412c691ae3f8de42d300"}
{"id": "train_074014", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "fc6dcc9a955fe69dc9a746bf5231292824afca8f"}
{"id": "train_074015", "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\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "c5302d93bd823bf0a532f00c25723f6c9bf61a0e"}
{"id": "train_074016", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "768fa6bcfb11043314a78d8a10c75ae5e4ad9eed"}
{"id": "train_074017", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "296a50a57650c5fb4ef3c82abdc814fa01208e5d"}
{"id": "train_074018", "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\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "244b32b108d5f9bb0e7baa73c8e83989a64b81d8"}
{"id": "train_074019", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "524b0d0e497d0f95b6ff534e164c7e658b8ac428"}
{"id": "train_074020", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "6f2bf30f7ae5c6576a70004cdef22513a5a92a15"}
{"id": "train_074021", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "f25101a2a6cf7c32177dcf6ed0c7e164115b3eea"}
{"id": "train_074022", "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\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e85f4f0725cfb1ae5182d02a474f1f1b16d0b4e3"}
{"id": "train_074023", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "5e0f4231d1d2232a6c1f8ac70017fe4ca40700f8"}
{"id": "train_074024", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "8404805c3d317276d6344fea20427cc34bebebd6"}
{"id": "train_074025", "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\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "c84aac87782ec8e8cfe2bcf4c8ff46b90e676178"}
{"id": "train_074026", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "2a5e6a2bab4b71f95e08c713647c506ef5c6d2d7"}
{"id": "train_074027", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "678ae6324e32b2ea92e1005605a87384f785fa4e"}
{"id": "train_074028", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "41dbc35381e4c57d8efbfc9f801f312d48d1e3b3"}
{"id": "train_074029", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "d1d44ec39684f68471b1f427f17d6934cb75babd"}
{"id": "train_074030", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "3b3afcb9e21f34f46c5aee759b7a4987917528a9"}
{"id": "train_074031", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "9feb1dc7a2376004e74102b112674e3db2903e54"}
{"id": "train_074032", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "0895105654a88c625038ec3e77708a8331b295d7"}
{"id": "train_074033", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a62b270e25268ebc9fe750cafc38886931d73195"}
{"id": "train_074034", "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\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["performance", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "9dc8a7702e949bcab039f18323ab5bf69b3f9c0b"}
{"id": "train_074035", "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\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "327f54c4ed82a9cb593e4ece5bc0841212e605c5"}
{"id": "train_074036", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "5eeac02ea4f47ed93f8243945c77c081c2b65025"}
{"id": "train_074037", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "85f953d005c49ecdaa2b22b13cd6b2e0b1ac93fa"}
{"id": "train_074038", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "622a8f1092579e5c8638a1a05d1be5b6ed499905"}
{"id": "train_074039", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "d785e5af77c109a8988c68393804970fe462929d"}
{"id": "train_074040", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "3c29c3a0641169ee0d7d3b65e106a24f1f974aa6"}
{"id": "train_074041", "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\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "e947919f28bdb83c8f37710ca5a117efec9e7e89"}
{"id": "train_074042", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "7767ce1fd0fe6356e8752d1467484811e63972bd"}
{"id": "train_074043", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "4c4b66914abfdd732bcf425a0af1021cba40520c"}
{"id": "train_074044", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "0d11d478a7e80299985330722f403cb7b718a225"}
{"id": "train_074045", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "b47cba1b49b1e469d36cfd2c0b019a180e37f2cb"}
{"id": "train_074046", "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\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "444c87e3f0fe13a5a3d34266be8d7f4dc8c738a4"}
{"id": "train_074047", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "5d99586bc06bd81208c069c417999dc173645e1d"}
{"id": "train_074048", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "8de3f3861ba4e2d0fcf9b0285768e78927ac7b48"}
{"id": "train_074049", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "074a2da92ab79de368663c46a5a1c6751ea9175b"}
{"id": "train_074050", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "be929c8da007a522a84f5f04f0f03e9cb8c48209"}
{"id": "train_074051", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "1073c3e2a533755b9f7dc5c7eea93a622cea586b"}
{"id": "train_074052", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "4482def31e74a77b1af177836b16279f4bad120d"}
{"id": "train_074053", "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\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "e725c96b376c7af6956d9679968866f4f961f026"}
{"id": "train_074054", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "4f040665044a4826e4934c4059aef70b0c743f13"}
{"id": "train_074055", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "cef98ee25a9fc7c40cad115659dd4e4bd29e6661"}
{"id": "train_074056", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "5d1a22561aca755b6c69de08c3bc84bc8f507e1b"}
{"id": "train_074057", "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\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "9a63a0ae61b9966f06740d23f08a8dcca06e1ba8"}
{"id": "train_074058", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "fc241a8c12a2382a2c9b2283b5c347c7507d049b"}
{"id": "train_074059", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "4f1705d3360c2d2209b300c11de919e3a4b5051e"}
{"id": "train_074060", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "c3065e440fc7d39e641f6f7c0bf6a5ad1171669f"}
{"id": "train_074061", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "8d783ac9e61bf36813e3618d77a289418327b976"}
{"id": "train_074062", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "23b328041ee838622a9379d41a1cca9b4a7a8ccf"}
{"id": "train_074063", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "0d2f3b8f764b6c7110d0387c3a50cff9fc43ec60"}
{"id": "train_074064", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "ce64062c878b9d24d78f52bea6720257e1f89d79"}
{"id": "train_074065", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "00d0f2772cf9bc7d77051bcb7453ad486ef7edf2"}
{"id": "train_074066", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "3b68b41d5c15b0a13b248c9d8eabbd52383ea6a0"}
{"id": "train_074067", "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\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "22336b076bc207d0d419ab48759c3e387e26ae4b"}
{"id": "train_074068", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "dece8d25bde01c5637b906d788b3afa4ab606c33"}
{"id": "train_074069", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "3b8ded8b4c0af6763a7ad2ea1daecf5eb6a6b1b8"}
{"id": "train_074070", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "da6bf57eee700dd5e4a137568241d6ffc02462b4"}
{"id": "train_074071", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "54bbd19f1c95480ec25b66fec41819b4d8da32ae"}
{"id": "train_074072", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "e79e8e3ed182a605d6b78f619cae9405958642c7"}
{"id": "train_074073", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "cd561c252f5e9b9006bbccd8f89aa231f12e1a9f"}
{"id": "train_074074", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "f3249e27035feb5c9088d49e4b3ea072c52ad23a"}
{"id": "train_074075", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "cf5a2cc375594a1bf71436e793095c5c0524f9b8"}
{"id": "train_074076", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "f36886ea02864b9bff5cc7afa75c60cccf78ec3b"}
{"id": "train_074077", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "3f5fda589bf383a7942d0be4101bd72db4f84326"}
{"id": "train_074078", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "9411fc809d078e3894e3b1ae1cf088aef5eb52f2"}
{"id": "train_074079", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "aa117564f0596fe67f576587cb774eff42b32050"}
{"id": "train_074080", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "5bbc5bd10640d01d75cb386f49c51902788ab640"}
{"id": "train_074081", "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\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e699fdf6c88d4b91fb3f67152c72bdeef0e08819"}
{"id": "train_074082", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["research", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "e38ddaa8a8cadeb54668087a9aea9b84c1b8b0b4"}
{"id": "train_074083", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "a995bad63fc6c8d2d402a993ea3130e63cdeec49"}
{"id": "train_074084", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "c8531e5bb07c478426535a667c6ad512b616210b"}
{"id": "train_074085", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "506a5ce721a04aca9a74af5afb30c2a41761d94e"}
{"id": "train_074086", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "84b55f9581397b7f77ca5cd465d89b48ddb8507a"}
{"id": "train_074087", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "9399a4c3e11d84fd64544841413a20505644f5ae"}
{"id": "train_074088", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "ae245f966cc6c760c7c3e142055da22a34bee049"}
{"id": "train_074089", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "30d5a4aa44735b8c403d0e50d0a279172e392aab"}
{"id": "train_074090", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "4427c76040c63ae5859e1442f0b9fe3f2c1012b9"}
{"id": "train_074091", "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\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "44837270b443e2017473c0e275d405f10c3750f0"}
{"id": "train_074092", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "99dc4cf3bdcc79b45bfa5f16cb32f3bbe76d91a9"}
{"id": "train_074093", "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\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "506a757c410b90fd6180a94b98e6e13b7fa1896d"}
{"id": "train_074094", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "70df8f5d457b71f9883ce263125351760076f93e"}
{"id": "train_074095", "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\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "bba212f49d8181d0141ed10ef46ffaa3f11a0533"}
{"id": "train_074096", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["agent", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "31a26834740a7591fbff3cea9e2838a111d7d3c8"}
{"id": "train_074097", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "4f2c3d4aadc16e048d4de88471b72743b6821d2b"}
{"id": "train_074098", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["performance", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "e9c61b5e62d470154670488108ed77e83e8d4e71"}
{"id": "train_074099", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "0af75ca2137d8980662343bc56ebd095fc55de54"}
{"id": "train_074100", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "f1179b7b3ff9eca6cefab9c4325d85b2d390cf16"}
{"id": "train_074101", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "643b6fc717244e138a6492f76a7c1fb70268ad53"}
{"id": "train_074102", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "61e78ef10eae03ec33575f117bbd3a46c563f3ff"}
{"id": "train_074103", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "71e88f9946e09f16955c3f0e3708cb5a52c89258"}
{"id": "train_074104", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "6b439307e49113b836e72d1429f50837336d0d76"}
{"id": "train_074105", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b27318f2d3abc51d92868f89b64f6f48745566bc"}
{"id": "train_074106", "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\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "2744499457e21f5d3666b4569333ce9226dd2b25"}
{"id": "train_074107", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "0936adefc23ea299b1bc5732af978aaf223312bf"}
{"id": "train_074108", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "c7c213df977d627fc1238781d6831664bf848e7d"}
{"id": "train_074109", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "f5829d6f3560219625772eeed4f0c57de2724908"}
{"id": "train_074110", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "9d6d6415d4daa4df16f5d77c6ceff735ab9302ca"}
{"id": "train_074111", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "eb33ade524f3b39e81364cef2fa07b037f87eaef"}
{"id": "train_074112", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "08b2f3af640ecff2aa32b6b50aa00b307ddd7ba2"}
{"id": "train_074113", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "11dc415f77e42160a724167e5d5573b5408e574d"}
{"id": "train_074114", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "55fc7b10fa0aa8a732d6be59d4d5f8412ca35734"}
{"id": "train_074115", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "8157cfeeeb35acc615c43e618242d9ed08ffa6ad"}
{"id": "train_074116", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "c72f18c5eea17a9327ad345e58c9e998f5dbe84e"}
{"id": "train_074117", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "85af3b6947a960ef7aa5b545c3f052847fd2ff08"}
{"id": "train_074118", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "1884ea66f181918d66da06bbd9611e237918d6eb"}
{"id": "train_074119", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "ef6c0df6bf76c24f4de879f28e258847d860ebf2"}
{"id": "train_074120", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["research", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "a42b138983a384ea4ef9327b4b687732eb2b9440"}
{"id": "train_074121", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "5f3b7097018b1093e53ca7b9aba5088c0383d33b"}
{"id": "train_074122", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "2f7ec88138d2b709b9d0c9ace328cb8a05cd04ef"}
{"id": "train_074123", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "b514d3b2e98e5e07715f3d2f8bc07ab64bf98a6a"}
{"id": "train_074124", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "3c38858378ce2b6af046d2cbbe8a1bcfdab0493b"}
{"id": "train_074125", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "b6278d88bffd05fc2b722c3b554c79a1b87099e7"}
{"id": "train_074126", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "5200801a6b22b8a4ba7813b1b65374fd229886d2"}
{"id": "train_074127", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e3739064187feba6cba408f76df56fbeb77272b9"}
{"id": "train_074128", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "822b51e748e41016ca12f05d9e182553579ceed5"}
{"id": "train_074129", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "36ae91617ca72d828427514769cd69724fa881e9"}
{"id": "train_074130", "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\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "b5ba44862c5a40ea16f0bd14f106ea3e07bee0ab"}
{"id": "train_074131", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "406c18d0f9b22f1b50a518821c3dad51749ef2ef"}
{"id": "train_074132", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "8be4de57c505b41e69e1e7e06bf7c51725d1668d"}
{"id": "train_074133", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "397de0a0711d4e9569a151bfad4cc6e32b383f88"}
{"id": "train_074134", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "3b2859629fb805ca48effdd52b323dc98a83fa7e"}
{"id": "train_074135", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "7e8d4b070707aad57cd4348c18aecf85de27e9a2"}
{"id": "train_074136", "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\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "fa4d8b25e297d0ad2030583d31057c24a4489e9a"}
{"id": "train_074137", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "3c068f4dd88cd7e394cec2228f4e3cb2d7c5caec"}
{"id": "train_074138", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "cbc50b5e45e05128a9b204dddeba54de0bb6b12a"}
{"id": "train_074139", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "9c7ee9a176c4f0c477928e6592500935cf0795dd"}
{"id": "train_074140", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "818c46d8a98b8d6f7e16aac6e7a608b1431f20a1"}
{"id": "train_074141", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "6c94b8306633c90aaf43cdcbed9d73eb78bf8a3b"}
{"id": "train_074142", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "dac48c20072c4dd511dbd4aa086c779c6cda3450"}
{"id": "train_074143", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "0a30c09eef5ed88f67cd82ebb118b359873ef8ce"}
{"id": "train_074144", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "0b29cd080c6a6900437d7b94377108b3a8c89976"}
{"id": "train_074145", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "757bd0c7503cc5dc9a2904549710ab565c905d42"}
{"id": "train_074146", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "26926dbed6352d1f4fa2afb69c1d359956476504"}
{"id": "train_074147", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "52834b1f3eecaa1ff1f6044517b17679b4b97dd1"}
{"id": "train_074148", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "5424b18563f2f1e7f7212c491d48e83785397abd"}
{"id": "train_074149", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["agent", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "7da8ca702e988d7ba4ee703b06f4256b367f728b"}
{"id": "train_074150", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "3789282035271e8ed43ee147c345ca87481a4988"}
{"id": "train_074151", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "0e0b34c725f0d6276faa6b1dd14f73cf0ae46348"}
{"id": "train_074152", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "85558d4201e13e92ddd44089053da4cdf91d91dd"}
{"id": "train_074153", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["coding", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "28881d5e67c93960837b7529affbf70057378c52"}
{"id": "train_074154", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "3d11f80f1033b48fe42f83e271d42464273cd724"}
{"id": "train_074155", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "5efc1fdd93bde50b36b3dfc0fbe9a84251ac4ae2"}
{"id": "train_074156", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "3582ac10866c50b1ea8fb5c7452c1f2eb3f7fbc1"}
{"id": "train_074157", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "18cc83299418af2df6a6ab265a8664b270bbcfe1"}
{"id": "train_074158", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "1ebb4110e0217fa0de9cfa198a03d22c4c117775"}
{"id": "train_074159", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "42bca4fac3a21d431a901837124f395c1754f2e8"}
{"id": "train_074160", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "a9b5bf740d9b40e1ab5aebc447e9b743f83f9220"}
{"id": "train_074161", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "b5a845b168341b9643b6a0766c9484bb7e22a537"}
{"id": "train_074162", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "701b6f837e99243cec33f1237ccf7c1847f5a3ec"}
{"id": "train_074163", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "537358187388b4279376312536a1befde6b35a41"}
{"id": "train_074164", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "b3d46694475e9485d77428e05e75b2b56716e5c1"}
{"id": "train_074165", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "5b52457d4e4f428696c098f41806989d406a731c"}
{"id": "train_074166", "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\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "e462fba52c0d3b3b543ef5351cc21a0e0c552d36"}
{"id": "train_074167", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "e9943bedea11c8d5922edff90d986ae3afede6eb"}
{"id": "train_074168", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "50f1f7afe13d8978c3ef3c674394287c41f1fd45"}
{"id": "train_074169", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "381d0c19fcedb55e388daca4de5d3ac2ea1626cc"}
{"id": "train_074170", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "2360dcd4f224fb7379f9c229d1c176b3eacde44e"}
{"id": "train_074171", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "21091051c09e1ea4d80bb2a99f159647b8b8f0a1"}
{"id": "train_074172", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "2a711daeff439f98d87ed5513aac64c55535903e"}
{"id": "train_074173", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "ace43f051378a6f1328eeff9ff4ab4470891f222"}
{"id": "train_074174", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "374aab7b53f099ba7ba0de0e81d2dfdc71b908b9"}
{"id": "train_074175", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "eb36ba23ffce6b72378e07cc69df7b91e45dd44b"}
{"id": "train_074176", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "63d6a5183fe9568b70c2b9349d8324e9c3072504"}
{"id": "train_074177", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "25119ba111d3dd1a0778a3939f8d60d8f91315c6"}
{"id": "train_074178", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "212f8a1aa90684e9de7f3fa3287e3fa75b1d515c"}
{"id": "train_074179", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["security", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "368b5ee330914aa22a200a7a49ed42e0e4468b64"}
{"id": "train_074180", "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\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "b33ba2dea4fd68642d862b731b93d774af42e8c0"}
{"id": "train_074181", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c91272c64b6ab27110dde263b09c3bdc05d3e0f2"}
{"id": "train_074182", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "897f9657b5e5f7af94d197755ba53a68c8bec7b2"}
{"id": "train_074183", "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\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "f4be643f6ed2f66456ba0a593fef794fd3e0cda3"}
{"id": "train_074184", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "dec5fbf2d07f86149230108d072beed0ed6367ec"}
{"id": "train_074185", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "279bda2fe42e931345b400e0e49c3307a8683f06"}
{"id": "train_074186", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "711f4799ff1519d217c9be0062ef1124ac6acc0f"}
{"id": "train_074187", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["governance", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "9fb043d814be4fb85c00ece288cdc9f0abc1fbe5"}
{"id": "train_074188", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "88e95c6c9e2411b551cda6325d308fee7e53652a"}
{"id": "train_074189", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "f0e223698f55a3d9c8d29a0c6d98d6ad52371f0a"}
{"id": "train_074190", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "04cf721dba3a53c99f5dfa1226faa1b3194bb324"}
{"id": "train_074191", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "42aac573169f422e024fe4cd8a8602a7b145472b"}
{"id": "train_074192", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["governance", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "d9f3cea41eba3e92a9cc3f406391828e8ee57051"}
{"id": "train_074193", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "5098cabae6575fd748f6160bedde1548d5429e22"}
{"id": "train_074194", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "75066fa7f3e25e093f5a2f9a50699701b2d995b9"}
{"id": "train_074195", "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\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "ac23099285c1eeb8d2369ebddd8eb39d522f632b"}
{"id": "train_074196", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "4f7e800b6b272fbb4c7a34c0a4ff9916e01b33eb"}
{"id": "train_074197", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "839844a7d7c735d72b90c406b2881d15cbdd89a4"}
{"id": "train_074198", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "76731075d4eb4e5b69d996eb6f85f8787280b96e"}
{"id": "train_074199", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "bc319dcc138471c8dba646c25d7fce7c1f01218c"}
{"id": "train_074200", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "144379d37767c9b0c8b609782e359489f634b7ce"}
{"id": "train_074201", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "20664917d694702cbac088e52e1cbeeb02534b1f"}
{"id": "train_074202", "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\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "9b55b285de03f3b715f67abc1d1f496d95fd5e71"}
{"id": "train_074203", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "9e18722d32063130a4d24fc88669233e1c551682"}
{"id": "train_074204", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "af4ea35e355aa9f030030bf170832e82c192c038"}
{"id": "train_074205", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "1b4da5c481ac1fe414b0671561937a0ffdb3955d"}
{"id": "train_074206", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "f640b90c36bc45914db3165d7a05933d6c97a242"}
{"id": "train_074207", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "61fd9fb099ceaa90472ed1057afc8080203adb37"}
{"id": "train_074208", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["performance", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "0741513c7d4ecfd955fde6800aa70eb6665002b4"}
{"id": "train_074209", "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\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "1cde996d8a4ac5f366b6978db318cecc440e0a2b"}
{"id": "train_074210", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "546f722806d7a002127b380905c3f3b21615c6c2"}
{"id": "train_074211", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "049f3dfc70a089b7d2750147e3fb845e7b482d3c"}
{"id": "train_074212", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "9abf4741225457eecaeb76a79978546f0d4dadfe"}
{"id": "train_074213", "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\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "4c07a8f3ec0b285f6d7ffbceb9f62a85e3457dd6"}
{"id": "train_074214", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "a45446a4ad3df2fdc2f42992de3c36016aa235dd"}
{"id": "train_074215", "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\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "58c816fa1c972ab11fd1a6e924a234b84fdf9483"}
{"id": "train_074216", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "967becdc58aaad9e893b60d52cc1be5e695fdf16"}
{"id": "train_074217", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "8c582d5f1c0c710a5ab5a301f787adeb396a7755"}
{"id": "train_074218", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "4d2e532b4913d9c7d1ed88d421b5bae265c3e6e5"}
{"id": "train_074219", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["performance", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "b07819d0e1463bede1be2b7b4a469127cb364278"}
{"id": "train_074220", "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\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "9f2b7edd9f3088877f964e48433d0f54ff4eedf9"}
{"id": "train_074221", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "532fb31587fa2c8c2f00ccd6bc691e8b52568a7e"}
{"id": "train_074222", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "bc6672aafbff49700d6b64a4bdd37c41b4d31075"}
{"id": "train_074223", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "288824953c7f913583b918a9892132f639d8c835"}
{"id": "train_074224", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "fa165afebb564425a6071b2d81c5fc3b38ddf133"}
{"id": "train_074225", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "2da75714e77a99243961151635c0bd6ada7f0f13"}
{"id": "train_074226", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a40ba009b78d20e748583411f33b1ac4fd6904ae"}
{"id": "train_074227", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["agent", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "e4f2e6e90b842844879f6c7c013af48f75bee039"}
{"id": "train_074228", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "462d9b154e12a6508aa1008ab3e2a4fd375fbcfb"}
{"id": "train_074229", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c0c228940ff7a30dedb628fef39ee683132e1f51"}
{"id": "train_074230", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "af800bcc9d57cefc93c725f95d44571b741a9105"}
{"id": "train_074231", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "0c07b48d7f9c0247f915a72ddfba4e0a60f1ca6b"}
{"id": "train_074232", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "627976cc6be57eeaf309f9de2d9c3577f03e0229"}
{"id": "train_074233", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "f1a96f39ba88c1934fe8864528a0fa1f38aa6aaf"}
{"id": "train_074234", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "7f0ab794e58fb04df248819d2c16ea6134558146"}
{"id": "train_074235", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "0fe37ba9c09a0a8059494f877f395ada7e478dee"}
{"id": "train_074236", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "67bfeb5bf076ca579530245505e80a1d3469ef92"}
{"id": "train_074237", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "abcbbf2c0429ffa3094f394e2589e58dd1c0ab7c"}
{"id": "train_074238", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "8593a841be041cb09388b489182b1e8887d42bca"}
{"id": "train_074239", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "81ab9745f764b1e3979fca5fa63a178be84d1aff"}
{"id": "train_074240", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "e741dfcc7880869bf6f8c45d953e5eea6e806ca3"}
{"id": "train_074241", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "b66170c152182680ed647e703e8df7ecc2c9e26d"}
{"id": "train_074242", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "cc2acf8b8e7949dabaff5258fdd3394449834fb3"}
{"id": "train_074243", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "9ca245306e3b4f99b42ce81da630672132d8d4f0"}
{"id": "train_074244", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "0bb10d23c22daa7b5b300ff34c94c8b583187344"}
{"id": "train_074245", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "25ca219fd813a5658d5c61d2afea3f0a092e4c22"}
{"id": "train_074246", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "c740ed73a689c3d127e65f6e87050ea1d1d5cfe0"}
{"id": "train_074247", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "2856ff81eba8c33364f3d2604289e9fa41aa3c90"}
{"id": "train_074248", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "651204f672767f27902f5d866b43942dd474477f"}
{"id": "train_074249", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "28298afb975c3a68639a86e9804c490c03905286"}
{"id": "train_074250", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "11522dda8aae6e31bdb3ccfbde0797a48e1090a0"}
{"id": "train_074251", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "6f02359e139ef17e66177b22350b8332c27c2396"}
{"id": "train_074252", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "4c3d030ee7abd59074fbd3484a00d445efdf9756"}
{"id": "train_074253", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "fb1e5c60d019dbcc9d6dc19cc4f3f5a0c66cce4f"}
{"id": "train_074254", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "c8e3263567e3bf6deb08a8bf9fae6636209937ca"}
{"id": "train_074255", "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\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "3862ddef5dc8ce80c36ed4d9122dc54c7a02ab1b"}
{"id": "train_074256", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "1fb1890959360671b3ef781cde7dc97a565599da"}
{"id": "train_074257", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "63c261546ae65e2e86f62f654e273e1ca2d30aa4"}
{"id": "train_074258", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "fdbbe882557f1f4b194ba9b667c66f1021a473f4"}
{"id": "train_074259", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "2e34991ba666bd2edbb91351ae0831ba5e5921de"}
{"id": "train_074260", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "178baa7f5a245ca8274660353036422205f21f51"}
{"id": "train_074261", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "a7d6707901c0e473589496f2b0fc8428a33f01bb"}
{"id": "train_074262", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "c012639704990f535e8727240b5876451b403b38"}
{"id": "train_074263", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "bead31dad8c126dbe8fdc4329b101e281065b3f8"}
{"id": "train_074264", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a3544e72aa24f056321ee6590387667f21a38080"}
{"id": "train_074265", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "e777da60e3b1e6491d0494835b3deb605b266b12"}
{"id": "train_074266", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "65ffbf2208afe9e075075d7b33fc7bb504d85644"}
{"id": "train_074267", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "919bc9c98f2f1566a09e62f79b1f3bf2e4590998"}
{"id": "train_074268", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "8e44a822d6e30b0c81df18fcb68e3bc9e9eae1ac"}
{"id": "train_074269", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "24ec2627a34db4077b53c6533927958f3d0b47c9"}
{"id": "train_074270", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "95fc4d74b1542d1b1a4741dbbbd8c1bc9e2b578d"}
{"id": "train_074271", "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\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4369397aef07c4cad1c6eb49058d373f0ecbd391"}
{"id": "train_074272", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "d0c9dc7ee15bff60d6b822501aa34d0f6b789ab1"}
{"id": "train_074273", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "3c103bf558769fbdd679def2f2280458134240f8"}
{"id": "train_074274", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "f0f7d26e4750ab63599b4c96fb448521af113e73"}
{"id": "train_074275", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "f6e4a72098a6114941ab4e786947db94453edd69"}
{"id": "train_074276", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "c383dd2b1da123bbcf96f8c3fab6db951293e62c"}
{"id": "train_074277", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "e2560db4f15202bb0b0cc6e7e15fe59661c98617"}
{"id": "train_074278", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "7c13398cec1d09748bcd17faada4299581fcb87b"}
{"id": "train_074279", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "8a1a5c546ec2a4f5ee27fdd46d37e39eee2d9e9f"}
{"id": "train_074280", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "fc639d8016ad5814be93d1dbca93de8113658519"}
{"id": "train_074281", "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\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "6cdcefbe108298c3adf26fcab01d1bbaaac67493"}
{"id": "train_074282", "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\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "2984d5a866468404d0861a73d8b6e8b8e3de0ba7"}
{"id": "train_074283", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["research", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b4119fc3b1b0f048a7cf9d960266f5d243e2971c"}
{"id": "train_074284", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "5f43a2d864b3109a5bd7fee88596227d201478e9"}
{"id": "train_074285", "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\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "b5ba44862c5a40ea16f0bd14f106ea3e07bee0ab"}
{"id": "train_074286", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "3fcfb570a349e19b0419c3c32faeb7d8d0c2f7a3"}
{"id": "train_074287", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "af683eca50879e67c6ac8be6ae6578fda552d566"}
{"id": "train_074288", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "bcb70d7c8b12cec568da4235b75be23d453a54c0"}
{"id": "train_074289", "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\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "80487feef9b272b7f8aac9330a068c2c2435addb"}
{"id": "train_074290", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "892ab30006849b5c8fc8beb23b4da015d2dab887"}
{"id": "train_074291", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "9f3a96a5649e09801b262ab0cd132653d072f3f6"}
{"id": "train_074292", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "0700dce5015ef3f8f9b48604554ead2e16c9a4de"}
{"id": "train_074293", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "8f90cc6e45fd065f3872f7dad0c45221589323b3"}
{"id": "train_074294", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "62e86d2da29f02c81bace762f26ae8e5bc73812d"}
{"id": "train_074295", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "2edd379c63616f0c8f7cfc638d49c4dace60711f"}
{"id": "train_074296", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "f8c0930a9e2eb49f62a1ef61bb69fba9d0836211"}
{"id": "train_074297", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "9d155b22dfdd2eec662fa600ffe7e6d485928799"}
{"id": "train_074298", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "b35a9c716db24a5a8a50919e10a0c6b731cf5a78"}
{"id": "train_074299", "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\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "2159802635f317f29013dc23e6096ffd2796b733"}
{"id": "train_074300", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "16a8323b977f54768bdb91e162a72edc8d50bf2b"}
{"id": "train_074301", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "032756633a9ff6108bd44e3f1daee9ff2448b5ff"}
{"id": "train_074302", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "0a9d6cb8cc119c91b1aeaaee64388023672c8fe4"}
{"id": "train_074303", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "cb69aa3c4ba85e61726741b98b3ea0689ef3cc6f"}
{"id": "train_074304", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "8798f178322fc2f4184fe027ee8bffef5a61d170"}
{"id": "train_074305", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "d9cc9584947052604b76ea47cf862b94752e1adf"}
{"id": "train_074306", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "8a1aea630ef7add9ca297e4fb00dd6320cd702da"}
{"id": "train_074307", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "33ef76dfa41b4a0ce62b95cd3c9f40079640eed0"}
{"id": "train_074308", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["evaluation", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "a084deca0468794fe11e04c0ec15788dcf99e40d"}
{"id": "train_074309", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "831b6662ca9e2d9f020bf48d8c14dd66a78c4ed5"}
{"id": "train_074310", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "5f840a395a46f7c63b92f7c3aa88f817ace11e1f"}
{"id": "train_074311", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "2fa949c7ccaebd598cd5afc3359179165decd535"}
{"id": "train_074312", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "2535b5869042407e38e9f1fe79409b5d8e72ef29"}
{"id": "train_074313", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "4dd0109fe6d2940e1f03d29af2ebb0ac2b04e999"}
{"id": "train_074314", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "62193287615ea4c1c4061a9ffd6be3d39a3a4107"}
{"id": "train_074315", "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\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "9713d037423ed7770817da31bada46bac0bc551d"}
{"id": "train_074316", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "81ab849d1e608a509c358a854a8c7aa78c8bdd18"}
{"id": "train_074317", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "8e2b17f448d2763a048d5893b564487496c1c96e"}
{"id": "train_074318", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e1933f09e4b93316aabcfd8ee4b5342d39cc09ba"}
{"id": "train_074319", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "3ee8883795960430c3db5421ced93325256d336b"}
{"id": "train_074320", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "80754e7d9296271897091f6168b673a66e5b15d9"}
{"id": "train_074321", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "7587468adeaf9dec7e2c3c0c721fde1b18a4d017"}
{"id": "train_074322", "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\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "d667fbd991d9dac5268b585a8e4fa0d9b9876550"}
{"id": "train_074323", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "53cd86576024c677925bcf50b72a8e14b01325f1"}
{"id": "train_074324", "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\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "7c8f21470d5740754f01f25f82b09517ebf565f5"}
{"id": "train_074325", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["performance", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "01b16b75482c36e37468d97c217e91c6b9ea0634"}
{"id": "train_074326", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "e6d2ccd522fed1715919327a000f9954c909a05a"}
{"id": "train_074327", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["research", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "df74a9cc84473d1e2bb10c1685f8ed371690c5ac"}
{"id": "train_074328", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "973fcfe3eb04472a63ef49a897748eedddf72213"}
{"id": "train_074329", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "6e5ec996bba70410cb920f80d144490529b09489"}
{"id": "train_074330", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "44f05f944843e3d1cdc3420f1a98fe3dc101ae30"}
{"id": "train_074331", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "0d79316061e63ea30224bf85a2b7c51b67f18b95"}
{"id": "train_074332", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "05cde68f38d1a74e3d797339d5e43a61afa44355"}
{"id": "train_074333", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "0054564ebb660b3c5a2a1bcda5b1fbbea2505176"}
{"id": "train_074334", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f3270b8684c8f9cd39462e3eaf3b96dd09c2cb6a"}
{"id": "train_074335", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "10f4c9471101d7036081886460a77c47f1405fce"}
{"id": "train_074336", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "7c79a0ee144dddc42076f41648575e0bd84c360f"}
{"id": "train_074337", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4e67f1816aba76564c540a691543a6667f12734f"}
{"id": "train_074338", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "5ef24bb2509397538d056f4fe5b714cfb2a7b43b"}
{"id": "train_074339", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "9b4514bb2102056aa9631adb732eca1a1d80d458"}
{"id": "train_074340", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "b58a2cb496e0b4aaf1072a76baf60c3dae81c2fe"}
{"id": "train_074341", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "9d717bf0172102a31e9d2e711abd6d5293b67e77"}
{"id": "train_074342", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "ee271beb1352fe16c99c59411a23535e03101b4f"}
{"id": "train_074343", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "cb05d564f8b2a58cffb61f1c122c0776c2720946"}
{"id": "train_074344", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f3f029f1d487cfb8842bc7d4dca60d2f40fc6efb"}
{"id": "train_074345", "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\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "90daafdc1bca93b5184fd6ac210abcf3cb91aad5"}
{"id": "train_074346", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "748cc9644ef5615396c4b9800198ef75f7cf2c06"}
{"id": "train_074347", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "40073463c38fffcd41009a10183f8c8c94572296"}
{"id": "train_074348", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "bcf588a993143b0ec8121b7d2f4f6c64b18e84b5"}
{"id": "train_074349", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "92bb03900a7e2eb0c906682b909bc978b1fca6f7"}
{"id": "train_074350", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b75d3acd5a8fe167006098655e7fb38e29e38681"}
{"id": "train_074351", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "de4561d51679353c3141c977b100ecc5f6daa113"}
{"id": "train_074352", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "d228b8b4ce8c66f765ff7f251b1404cfdd2aa66a"}
{"id": "train_074353", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "17070f88b22c2df29d7ff514370a499bf435ebcd"}
{"id": "train_074354", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "1c0bb18311f6fd6e6399496b77b79e8f151077a8"}
{"id": "train_074355", "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\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "aca37766aebb342e35dccd863a9f42624a05b140"}
{"id": "train_074356", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "55fe2915d3c8dae25aa8ac059c681d5519fddccf"}
{"id": "train_074357", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "2117c4da758236600aae5bd37542b353887601bf"}
{"id": "train_074358", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "4481214dee5e6caa360e3b1fe8f225cc9b84d8ad"}
{"id": "train_074359", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "42d6cd43db389a38686b77b3619af5e26adeb256"}
{"id": "train_074360", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["security", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "cddf5145a9759fe4e5a55c24d54cb35594480d07"}
{"id": "train_074361", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "e53232e275d0bce16a5602983a1bbbfa8df4bd6c"}
{"id": "train_074362", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "4e54a99301029d1bfefa5eaae1240ff6f5fc0d67"}
{"id": "train_074363", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "06cb0381a80bdb48b512aa60593e78ac7d3c5888"}
{"id": "train_074364", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "d3dae690ab719422ac41656212f983ebbfd9b7b4"}
{"id": "train_074365", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "a759f45423117992b376a14ea93d611c5f298d3f"}
{"id": "train_074366", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["security", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "6dc8a7933fb793ecb6e8720a48780ac683bd4c5a"}
{"id": "train_074367", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "f082ff81a8456cf279ede37563b6f846e0acb549"}
{"id": "train_074368", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "e14f70e8cbed6e9d38767a020929511d1c8e4566"}
{"id": "train_074369", "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\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "7fdd44fed47989e8c45b3c0e7259b378bdca27b3"}
{"id": "train_074370", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "4a88319fc18ed129de3a0c266537e1799d0a16ed"}
{"id": "train_074371", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "c7c15200eb61887ec67019e37bd65e50f813b5d4"}
{"id": "train_074372", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "72598a5229596d90d294d61b66263d72672b3391"}
{"id": "train_074373", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e3e61ec2c22d4885f70166cf3a6b18383d1a92b8"}
{"id": "train_074374", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "dc806b4742589c1565dff0abfe0258441baf7b58"}
{"id": "train_074375", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "7cdf5c919065fdc553ab64e99a0df7409b6cc554"}
{"id": "train_074376", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "ed4cf7da4780ba1390b461af2a7a00ef72646d19"}
{"id": "train_074377", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "bed38597bbc7c9f4b57b6f59134142eb974d1612"}
{"id": "train_074378", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "489335b0806d7dadfafe9ff7dc3eba0302c8f09e"}
{"id": "train_074379", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["agent", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "3544ef472fe59b603dcc8e4459f7796af568218c"}
{"id": "train_074380", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "b0f5cc3074364367ba82bf23486ad81567d6ac77"}
{"id": "train_074381", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "0fa4130f31fd46556dd2edd612c4e235df9ccfdd"}
{"id": "train_074382", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "4ca0cc6de3ebf4539eb27a412362917276744479"}
{"id": "train_074383", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "1f224516a16eec82988242350036be4555426694"}
{"id": "train_074384", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "5f66f00f95459d1c6eceb39754ffef5a5fde5fa6"}
{"id": "train_074385", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "22ad3c27fdf0f6909871480016b060bb4d336231"}
{"id": "train_074386", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "84cb3befa7f9e4ac59ca95471b43508e8c2b72d7"}
{"id": "train_074387", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "ca4657b9ea66bda02f712529e5474bf83624f2a3"}
{"id": "train_074388", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "447326527ba2869e239cb9c6fd5c0e80431a7df0"}
{"id": "train_074389", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "3658d430cc0b167c7a15780769eb8dedba6b9776"}
{"id": "train_074390", "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\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "76b7324acdd2a4b83a8b89c0fcce11f8b0cc415e"}
{"id": "train_074391", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "5bd2e353c84cff45036a63f0ceea65532b820fa5"}
{"id": "train_074392", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "3614213b800472787f3815db6a95a4540b55a641"}
{"id": "train_074393", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "b87230113028bf5b397879a749c5f36ea8dadd50"}
{"id": "train_074394", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "3161097cc70b04bef9407dd343a71551e1dfaad3"}
{"id": "train_074395", "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\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "b8dd95467c58f8093929d87c431bd67abf4ec173"}
{"id": "train_074396", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "cebf9c6bd046c5b0668d73fb3b2cdf21bef35f67"}
{"id": "train_074397", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "00b87ebdf1d3304639a09f5f632d1909a5cf2306"}
{"id": "train_074398", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "57d5c02c0fbc19915dd90f0e17d570446b95249b"}
{"id": "train_074399", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "31506cb1d0f68317059ce232557d0a6718a8ed7e"}
{"id": "train_074400", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["performance", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "eb6c7d714ca3f40bcd274e0430c7871b6ea2f23d"}
{"id": "train_074401", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "0601d44909b4d7e6e59c6d99da119cf30255a6f7"}
{"id": "train_074402", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["coding", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "213ee894f08e47a51a2f42d182e331d33ef835fc"}
{"id": "train_074403", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "fdece1fcd2d38ce45372d27ae64b3b7101cfe39e"}
{"id": "train_074404", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "1c0be31b09a5d77e88f29d286ff7e900268ec7df"}
{"id": "train_074405", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "31c97952e22acfeb0895ffcac903c03f75f023d5"}
{"id": "train_074406", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "63d448acaebe1f7a8a406914cce79fe541216d79"}
{"id": "train_074407", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "367dbbf77f87e23d3615586371f777f850264d22"}
{"id": "train_074408", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "a46ef1b49bc587e228c1121b2fea73bee2f5c156"}
{"id": "train_074409", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "9714e4e7c826fcd34f35dd8406157c1b93f21b4d"}
{"id": "train_074410", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "39dec78c398880d764f30d3f0cae74d684682f30"}
{"id": "train_074411", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "f7c30f12e534e04e0ee40df73c7a2ce5ac871eba"}
{"id": "train_074412", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "7dfb39af7160f9c264a5b96ff6a6a53d09304546"}
{"id": "train_074413", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "7b63c669cae428cb8c9b9f1f8d8563680cf14ac2"}
{"id": "train_074414", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "3c0b07085272557e1446baba0c4c52874722457a"}
{"id": "train_074415", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "bad0e4163ff97333b765523605656c4d0ab5c83f"}
{"id": "train_074416", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "b28744a4c20fe7a8aae3a1752456827ec0acc1b8"}
{"id": "train_074417", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "fc4f0fbf55ca578218946e64bcc924428dd6c028"}
{"id": "train_074418", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "e185cc0919039e1a9c627b4c578ccea474430635"}
{"id": "train_074419", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "faf3a368ec906d01e74752a748f0ff32dd3a961b"}
{"id": "train_074420", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "ea716d28335c4861bbcce4ec36b8f4a9ed99fe23"}
{"id": "train_074421", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "92edf949c7f22cbc0fc3dd64d6bc7f5198b0c12b"}
{"id": "train_074422", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "c957f53621dbbc272ea991e2f1076235402949d2"}
{"id": "train_074423", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "a35454f4bda5a0dbd7ef9d4ad1d05ecdbd21e591"}
{"id": "train_074424", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "b5ad848b3a467552dc6801d7ba840a5dd3ff90ae"}
{"id": "train_074425", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "af3ec344e09530ae956ab75d7298a085447126a7"}
{"id": "train_074426", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "8f1e62aff4aa092abb7110df3896e0a52bd6306c"}
{"id": "train_074427", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "cafebeaa6674f42d97e3d2e78f03cb20ecbf3068"}
{"id": "train_074428", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "66346e0768d4032bb15fffa1f479d0d3d941996e"}
{"id": "train_074429", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "59f4d2cb1904db1ad49cfbfe8594ef63a8cf6fef"}
{"id": "train_074430", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "fa680f76e52bfe2d80ccc0a0cbe973fa30a43a21"}
{"id": "train_074431", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "3c43b2eaf2d034b2e7b3413a669bd8bfc7236a6e"}
{"id": "train_074432", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "014573266554a3ecc3d4e0f42d5b691788838e1e"}
{"id": "train_074433", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "eaa891fe111acb7c5d0a15bc53a3c509e3af6d99"}
{"id": "train_074434", "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\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "55be245127c25ae089d367edb076174319d7e76c"}
{"id": "train_074435", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "17974173bd1bd9dc73eafdf47990521403507703"}
{"id": "train_074436", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "ec4791bf1adfdea36d4d9557489676082387e6d4"}
{"id": "train_074437", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["coding", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e3cea6e1a79ddfffcf9a05994cb16e6997d4f265"}
{"id": "train_074438", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "d462be2b439fb8eeb9853d2c7d553dd701911351"}
{"id": "train_074439", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "e064b7ed72c1041ce2675a9ebab6d51c91fc7fe8"}
{"id": "train_074440", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "d9edc7d2cb3ac42ba38020647b56bac401868c5b"}
{"id": "train_074441", "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\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "71e8a89e355d431e904ecd4e271c7ec13f938296"}
{"id": "train_074442", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "cf44506cf60f5bef7d20f8c5321b6ca284b7553a"}
{"id": "train_074443", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "70b0945c8a9af31f06326c6fa488cba399f9b51e"}
{"id": "train_074444", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "c0c36dae969943be0b83f876da993b099090090d"}
{"id": "train_074445", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "8ecea634951b5e8cefd7f17cc795478c6f5687d8"}
{"id": "train_074446", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "86cc2f726fae292e3ab00687a57608c649a345be"}
{"id": "train_074447", "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\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "34f9ea6642298ffd41140a3ec34176d6e65cebac"}
{"id": "train_074448", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "98564065d328819074191e11af5106eb4b6a2136"}
{"id": "train_074449", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["agent", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "64e5998824ec9c1105f4dd4c6e4d891ae4a83517"}
{"id": "train_074450", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "4f63ab84d60b494110dc32f3f155b71a7f24fde7"}
{"id": "train_074451", "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\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "a12bab527715f9e1e7d4fd9bf84ac8931a339b11"}
{"id": "train_074452", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "39145ab3cda9e1d34e8f181352b5bb31f0939f2d"}
{"id": "train_074453", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "26853e1d77ac754f38df9de3bf9ad2f3f5813830"}
{"id": "train_074454", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "631507006309e1713aeef0aee72f52355ff5c7b3"}
{"id": "train_074455", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "518df62129561b5fe11513da8f6254be784d5f45"}
{"id": "train_074456", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "3683c7705e6951a0d973b8afbf02d5ce71cd7b71"}
{"id": "train_074457", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "c8b07e0cf62f13bf876f6f65be884c2e81957ae7"}
{"id": "train_074458", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "6d85899bf14a27e9cd577fa71d0f6c566033b368"}
{"id": "train_074459", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "d31230e65a13a118b483a807497518d4de6977ae"}
{"id": "train_074460", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "32c84ceb8d452c779ae9df96d11b7f410c3cc20d"}
{"id": "train_074461", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "0e8e7659cc9ce219e60db5bd0c7f9bae24f78c0d"}
{"id": "train_074462", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "608de79249dcb385ccfe203ee4bd2197be646b4d"}
{"id": "train_074463", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "d8197adfefacf7d4a285655e47242aff4f3cc049"}
{"id": "train_074464", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "42639df93cbb6ccc762fdb70399650d4002d5231"}
{"id": "train_074465", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "24747a32fb3b0b5810cd5fb4eb38ca81c8438a4a"}
{"id": "train_074466", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "6596cce51b075b3c7e0e51b364d955261b2c04a5"}
{"id": "train_074467", "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\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "2f2969c46e47a114470383daf7245542df60aac1"}
{"id": "train_074468", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "74f4b71256fed1fa7fa6d10ed48a36c755d25be2"}
{"id": "train_074469", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c6af865555e57c7b12b7dfb12844787d0d8caa8f"}
{"id": "train_074470", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "ffe58eadd1460a9061b48d18e858512ee7fba113"}
{"id": "train_074471", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "8907c41f67ec18dd49292ab809eddf0bb0739887"}
{"id": "train_074472", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "a68ea8b4e125dae98067c7491e9d8698a28cdb7b"}
{"id": "train_074473", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "51bf1cc2587254d7a8bd57596c60e291d256991f"}
{"id": "train_074474", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "27e12fdf9edd7ff3a9afdff1eb91f235fd7577ee"}
{"id": "train_074475", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e3e62c6e31344b54cb0405a8798d4e9b6dc5dce1"}
{"id": "train_074476", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "be70fe6db1686d69c026dc14fe33b229733fd51b"}
{"id": "train_074477", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "190f40592679954945b107cf85dcf66d56178f34"}
{"id": "train_074478", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "8234a44df918ec598809fd3614aebf1106ad47f4"}
{"id": "train_074479", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "30d7aff1ada2d5986daadd8a7e9466540904e00d"}
{"id": "train_074480", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "7df8bf4a50b3fa994cc944e3cff29a26cdb18ebe"}
{"id": "train_074481", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "d1f29069b3fba5e06fd9627e97da3681492f460b"}
{"id": "train_074482", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "8c9011817cb8e4016db6b1ea63b5ce2b2eee7dcf"}
{"id": "train_074483", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "978e1a3a94ef314fe91fbfe576c300728ce18d38"}
{"id": "train_074484", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "6e1acbc13a7f640dc7f89b86c06e4fc24926de0a"}
{"id": "train_074485", "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\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "e1d247b3d602ce364d793850e29f73981bde58b6"}
{"id": "train_074486", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "65545a6514b51d400727a600c79a40bcc0e7454f"}
{"id": "train_074487", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "0f0d951bd382cf4f9c9e1d28ac4e115392aea05d"}
{"id": "train_074488", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "340ae2157df4eadb60f49136d069623ff581c042"}
{"id": "train_074489", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "a599a56194b7a6f0b9e9c325911af84673d406b8"}
{"id": "train_074490", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "ebeee5977408824882e8458815de347644f5d310"}
{"id": "train_074491", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "993c1f43e6c21520868f9bdc9571b5c169542080"}
{"id": "train_074492", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "404f0bb8aeba66a014c3060fef83dcd550a33b23"}
{"id": "train_074493", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "5db4389c43f555fe825261ca79d0cc5a2ae32676"}
{"id": "train_074494", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "60e47b564d93015a05617c51308475242c2b9616"}
{"id": "train_074495", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "df1fdc2b58f00d0e8b3527fd0864d244e0878370"}
{"id": "train_074496", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "59c1622286921abee0027feddb1ec01b23cd9a8e"}
{"id": "train_074497", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "7872fdd9ae2cfd48b3f0206131eb506621558e89"}
{"id": "train_074498", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "d8c01d1463bc5100c4d068e5f32b6d989d545e4d"}
{"id": "train_074499", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "50f82dadf2428901072454dc2ce690e00d6a24d1"}
{"id": "train_074500", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "b6e6c279831dbad758876166717354308ff47918"}
{"id": "train_074501", "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\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["security", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "eabe69670a2c1c55c04cc9837f93a85befa013e7"}
{"id": "train_074502", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["governance", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a3421609a8fc4daa56a239fc47f4a6055bc862ca"}
{"id": "train_074503", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "ee16ce437ec55f73e1cf5244836851f754fb3f50"}
{"id": "train_074504", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "1cc7e239dce5cd41ec0efdca700c009bf861c058"}
{"id": "train_074505", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "0e078921e1e7c6bdcc5c3dd0ec49901580805a12"}
{"id": "train_074506", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "13ea882351b0a2d5c157efb0801a35e3e86c9631"}
{"id": "train_074507", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["agent", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "d909f294e63f77342ae778ecc20a1d3a20cc606c"}
{"id": "train_074508", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "d88b789f292cfab4e0a9ed6cf8e19ffc720065d8"}
{"id": "train_074509", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "fdb3905f93a5ca7b4b002ec0aebb334b02994757"}
{"id": "train_074510", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "d4f2c36a77cb6b0de9756287bf48d0cef4eb3826"}
{"id": "train_074511", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "50616df38e5e36c02dbb4aa90e285f5691d31a52"}
{"id": "train_074512", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "1294e817e637a69e322f59e0bde6401404a357fe"}
{"id": "train_074513", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "f986413330cba0f05099fdcc4f573cd8fcc5d34e"}
{"id": "train_074514", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "952e27758ed3400d1bc46c1400b941b642c30fb8"}
{"id": "train_074515", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "e8d4f6d634de0122d62d80f97e19b9f13345c564"}
{"id": "train_074516", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "6e46ee81bba69b2481f008661b00c1c1e4675fb6"}
{"id": "train_074517", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "45e514934bb50dfb5df7c5dec3a02702a9c4d896"}
{"id": "train_074518", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "dce167984f4d7f54b19837e0d43bfb9504e9f88d"}
{"id": "train_074519", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "8dbd9a251137edfc333b6b0303f1fc884ff485f5"}
{"id": "train_074520", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "e0e326255456e59c945e7f9f1b17bc0b11b5e3ba"}
{"id": "train_074521", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "061fcfe72aee4f01f845f0ab731895b0253846b2"}
{"id": "train_074522", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "fb19854d1e4cbd11a5b9c8c329c29d6f28f56c13"}
{"id": "train_074523", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "84443dcc294ce31d9eed74f5b488a398bfe99f5e"}
{"id": "train_074524", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "19e3f1630ed57b42233893334bede3589107fc49"}
{"id": "train_074525", "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\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "16cf1267b788fba85505b751de14603523400139"}
{"id": "train_074526", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "20bae42e9ff16d92cb1c1c5baeafb1a9c866f309"}
{"id": "train_074527", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "ceebe7197ca3775b51706f8e16bd31a3cc66b5ef"}
{"id": "train_074528", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "24ee0f0f187ef2bcfd5db6fd20181f281a386aec"}
{"id": "train_074529", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "2ad6dbfd7dc423b907f65ae0e99243116e68b563"}
{"id": "train_074530", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "33df3fde93e3e776e9adaf31d77e0546442971a6"}
{"id": "train_074531", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "2af7ef7549e1eb532dbe6d9843b700232b7fc3d4"}
{"id": "train_074532", "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\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "b30d480a1f0e83ac6d94e458c5d7f3dc73a5ff38"}
{"id": "train_074533", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "35e6d34fe8e2444863946517ff7d445f190b29c2"}
{"id": "train_074534", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "03cd4a256a075b71a6bc0a41cc578de672f7b815"}
{"id": "train_074535", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "61668d63f33f4aedb942686baa64a78b40e54df1"}
{"id": "train_074536", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "3113e5d80445c581af4cb9c85bb53de49e13bc41"}
{"id": "train_074537", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "3ecbcf0b2914e82287561a29f316f7eb37ca05f1"}
{"id": "train_074538", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "c8f5f75f43c48f85f3620c3b344f0236b1a3b5b1"}
{"id": "train_074539", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "c0e6a289bfb58b189c8ea9c33d2759d8e6a10362"}
{"id": "train_074540", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "6fcbb3784a0625492453347d0a7c2093968d4783"}
{"id": "train_074541", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "5626fa2fe5e15bea2692261ec4ede8b43f6cbdb8"}
{"id": "train_074542", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "3959520789d8dee89caee9c042938dfc015f96fe"}
{"id": "train_074543", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "7c1241e1561f216b4eff79ffbbaca4490cc78e32"}
{"id": "train_074544", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "2e6b23202b348213482f9de4866033fd0513d103"}
{"id": "train_074545", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e9cc55aed05a6fb94b12da27e32ae1faf95afd30"}
{"id": "train_074546", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "cb725a3f9203a7ec9766e512fe6a239da55e5100"}
{"id": "train_074547", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "519b58e6318c49f315018848bfffa4184c194f4a"}
{"id": "train_074548", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "25d6353adc35801f4992e43bb3a35ff10fa51922"}
{"id": "train_074549", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "c055ae94025c932558abd402aa340015142cba0e"}
{"id": "train_074550", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "1ba1cb6a0a0ab341833c2f503716c42ae9e85c35"}
{"id": "train_074551", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["agent", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "3a0c86c2a75fb51f6d1220fdb2929b4dc586aa0e"}
{"id": "train_074552", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "ebe4c2c9d6011a0bc26a0e7dde9486c9bfdf5cb7"}
{"id": "train_074553", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "fb8c76db5b6f857bc346c29a3d37f0abdcb7334b"}
{"id": "train_074554", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "9398efad2508dec954078fc940287838c59eadc3"}
{"id": "train_074555", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "dcedb7b62e62c914b6251ca4726f185fae496cf7"}
{"id": "train_074556", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "90c2e0d2190eaf31f4b74bd20db11e2447c01af3"}
{"id": "train_074557", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "cecc30569abb2edb7ed088d22e9ab8cb102c4f40"}
{"id": "train_074558", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "eca95541ae9886f67ba3c4a098581239d296fd0f"}
{"id": "train_074559", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f028cdc59cfb28c93fb6af194b954a91e7828fa2"}
{"id": "train_074560", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["security", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "872a8af2daae1185d36a81d62186c6c4c9da1314"}
{"id": "train_074561", "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\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "3c0c8b1928a27bfd629ac494d1f37c1f7476b12c"}
{"id": "train_074562", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "2ec29f8a21e1edbdf549cab42e1b87f033f4345c"}
{"id": "train_074563", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "0f8e9b13d4395d91450c3f29031f88cd6fcdd48a"}
{"id": "train_074564", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "c55049e7ba8f3ebc943f029ca7c26301e21af6a3"}
{"id": "train_074565", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f8b2752b00e0d7b1333cd4362c1dca6e6e0862b8"}
{"id": "train_074566", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "3393a949294a2f3894fb02be49824ba1163328da"}
{"id": "train_074567", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "27d53e0ea9e70527b2602cd95bbd22dcdb1664a3"}
{"id": "train_074568", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "3937eec7ee82437de3baadba455ab88fe370faa2"}
{"id": "train_074569", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "97f1bf9df1eafdbb9a4082b53bbfe185004f1027"}
{"id": "train_074570", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "13fc68e53c2edb04f78ec2a573c2a9343d7a92ca"}
{"id": "train_074571", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "771aa04e9c4b3371a824700c673e227279bf2f28"}
{"id": "train_074572", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "1a8395995c5873a5c214d30f1b64a994ee9c31a2"}
{"id": "train_074573", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "6182ffc87f7a0a3e1dd6421780b623e0d913e995"}
{"id": "train_074574", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "20802ba09e303a1edcc8a701cdbc874220caf724"}
{"id": "train_074575", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "2571590d066a2fd51196db0225981a9befe64fee"}
{"id": "train_074576", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "c3295256f367cdd8f15ba93dea199c6cd675b34f"}
{"id": "train_074577", "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\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "88dcf573526bbf25bbf206cb21286e99d2546028"}
{"id": "train_074578", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f839f746bc3064b046af2db6f13904e8cf29aa96"}
{"id": "train_074579", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "c05791bbc926e75c752f5436be861334f70ce93a"}
{"id": "train_074580", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "643dd91f09f8b50c9b9915f411cf3eb2928fde29"}
{"id": "train_074581", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "d6f2e459424d7d4335d5336ad44250bf5301c52d"}
{"id": "train_074582", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "7523d93b37ece021ca8668105d2e9c11226480ce"}
{"id": "train_074583", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "497e18da3c1520c1dcee3803a8802207e27a5d48"}
{"id": "train_074584", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "306b0345898e9249f7262c0b53d037803672c28e"}
{"id": "train_074585", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "586d7c37e8dba8693797cb2871c73d96b76202ea"}
{"id": "train_074586", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "945e8dc7e7b6e4b2bc772de6c2b3d2c21939f136"}
{"id": "train_074587", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "b94d98d1722d18c5f302b6435ddf023d961b3b58"}
{"id": "train_074588", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "217243d4eaebd4d94121055d99e63e20dedb9cc2"}
{"id": "train_074589", "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\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "abc74192df2607942debc07e7e9158683348c9cb"}
{"id": "train_074590", "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\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "d8f37a5e60f037147c4bf6f5e35e25f10a8c1f35"}
{"id": "train_074591", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "1d66e0b2c7eaf4c163b03746362a27d92ea1b047"}
{"id": "train_074592", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "27ee7dfd8898529956cf479cb976c7c2dc3ff93d"}
{"id": "train_074593", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "7a0a2936414baa22412c6c1d72dd306538d3a192"}
{"id": "train_074594", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "0d725daa9d729fb548ec00bb0fbbb90b03c64111"}
{"id": "train_074595", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "9b75ec8c153182951dfd21c277b9539dbd76d210"}
{"id": "train_074596", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "fddbc3c4e3874b8c074e2270766eccaf326b2abc"}
{"id": "train_074597", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "397fb668c8e1df0ab36a92fab80544325d1f7b84"}
{"id": "train_074598", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "d238340776354878be7e7bfe4f44f08b63ea80a4"}
{"id": "train_074599", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "7c991cf676e397308efb4a2459f68f99c7dc12fb"}
{"id": "train_074600", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "b2bcdd4fc0ab51b5b3908dec874e39497651fc09"}
{"id": "train_074601", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "e21505d844422d943b149b7d7f40ea54a715269b"}
{"id": "train_074602", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "ec8ae4355e38d39b82e6ca03358199771b84d488"}
{"id": "train_074603", "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\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "6f554af29d4d196d8fb6a69fe04e92c3137b61d3"}
{"id": "train_074604", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "603e4f32c17c7d38a7b15f8f0dea09fc1fd44c70"}
{"id": "train_074605", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "76c5f8a5d083d736ddc110f4fcb3962f99e320a4"}
{"id": "train_074606", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "76f8738c42aef89a2d1b6a43dea52456da72727f"}
{"id": "train_074607", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "681eb3fae676401c84592460bbe48034b84f6291"}
{"id": "train_074608", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "cdcfc11a5a454f82798c34b8f0f94db1687b51c5"}
{"id": "train_074609", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "d696c1e05804dd8e84f54eba188f0bc8b2e90e2a"}
{"id": "train_074610", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "672779d1c363f5a05e68211522d4143faeedb3ba"}
{"id": "train_074611", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "5cd03c6c69203740c3dcafffcd5151acecd9cd80"}
{"id": "train_074612", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "08199d6d4bc2e88f6bb3ebca70df9014a29e2f4b"}
{"id": "train_074613", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "82ac0115f51412de66b75931853218e28667e810"}
{"id": "train_074614", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "f989ffa6f44c844fce6fa28368d2f425855534d5"}
{"id": "train_074615", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "0f3fdd178f4f34ab1dd2d489aa0cfe7c657ae7bb"}
{"id": "train_074616", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "bd641840bbe78b33ffa920e67b1da896a5ca1a9f"}
{"id": "train_074617", "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\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a10f823d023cb39fd5d48cd770d54d72ebaf8309"}
{"id": "train_074618", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "f8cb26c6a403d92d0b2f96f5db0b03c4b1df4072"}
{"id": "train_074619", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "169e473b368190149ba9425fc4fa5097175a4dbd"}
{"id": "train_074620", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["research", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "8ad99ff4dd0a537ea43b3b99f437f03537ab3b40"}
{"id": "train_074621", "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\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "eca7f9f546402669909688bace8deb3eb599176c"}
{"id": "train_074622", "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\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "057e66004362b2e784ce637df79ce62245dd4069"}
{"id": "train_074623", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "4f1141212e555e632fe924eedf289984a8556b53"}
{"id": "train_074624", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "134df6a1264e4d9e7212d83016904d9330594531"}
{"id": "train_074625", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "c046bbb6963612c8cf49ddbcc3c5aa5d84bab40e"}
{"id": "train_074626", "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\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "36d325a04986084b2b905af83fa00017e8afb9d7"}
{"id": "train_074627", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "daa601f7aee3258bca0ccd531ec1abb9a63f8a37"}
{"id": "train_074628", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "1939a5e7d8b79a6e7ef694713b94b3096468d3d3"}
{"id": "train_074629", "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\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "4222d8f35d53af0df527229f9c3677f705a91f68"}
{"id": "train_074630", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "2eba61469bebd1e9fe2b2ce59e7cd98730b23b62"}
{"id": "train_074631", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "1aa86001e6c5c8310562148b3db2b5b396b92077"}
{"id": "train_074632", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "2bf5e05cef1aa1c5a3e27df4e9959deb4b301df2"}
{"id": "train_074633", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f3e97f36bf66e896275a14db80c947c5c9af011f"}
{"id": "train_074634", "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\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "4cda20be150e6157ee24e98c15fdc02844aaca08"}
{"id": "train_074635", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "2c251d8ce9a70176ec1c60d6d002607261490573"}
{"id": "train_074636", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "bf0413ac73e73085dd8f272b10e986a8b92a7807"}
{"id": "train_074637", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4bd7b07cdd09b11bf76fe083657b8ac596ac42a2"}
{"id": "train_074638", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "6abdb25ba8a0375b063cd3306780fc267855bc65"}
{"id": "train_074639", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "8842ce66d36ad89bf9fe8684430e26b9fe771a17"}
{"id": "train_074640", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "cac4c9009b6ab4a4e59a74c0a71b835c1cdb0bfd"}
{"id": "train_074641", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "acae3e828a78da1aea95049579de909d907f4094"}
{"id": "train_074642", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "bc66b7af47c038e1a732887447293702e735732e"}
{"id": "train_074643", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["coding", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "f6e9a8063a1e62b879c40cdb0e2d6f5c3d06d92a"}
{"id": "train_074644", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "4b5dbf01714a9bc4304ef8891704c3b432d619c3"}
{"id": "train_074645", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "31ee8fed58856d237e597edc990f69ea0afd21ba"}
{"id": "train_074646", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "0e940b30ec472c8bd2455ea96f1107e4264810e2"}
{"id": "train_074647", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "f30d16c4c6b23bec248b600c4057e5afa74383b4"}
{"id": "train_074648", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["performance", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f0ab9a9e3577b2cf7390b3ab251cef39064b182d"}
{"id": "train_074649", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "0a67f24bb224a5737b8cd9f0d428145617f6202a"}
{"id": "train_074650", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "5f60bd1f9b8ded5303eec07d0230d2372510b375"}
{"id": "train_074651", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "bbf1d210d0406f6d3372adc40aef6a06d501cfca"}
{"id": "train_074652", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "e2655c4077c07539fdc9ba1642f17b5f1357ca1b"}
{"id": "train_074653", "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\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "ede3a8bfd50fd9e889090e9a54969afd1ad68d63"}
{"id": "train_074654", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "92c3997aa33603c9af22e994668eac65fd6f6930"}
{"id": "train_074655", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "5e2abb525c66a163609a2589a5e8cf8bb0dfd7b2"}
{"id": "train_074656", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "8a8ddf954ad2c1c4886cb832e443a062bbefc032"}
{"id": "train_074657", "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\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "504b8a0741f6f7d8c7de8fd9a9de0edcb3012928"}
{"id": "train_074658", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "9c7ee9a176c4f0c477928e6592500935cf0795dd"}
{"id": "train_074659", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "1dbb3d4a0ffe8f2e9ecb52a71f7e9b18725280bc"}
{"id": "train_074660", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "7c4fec3d89ee584fbf021d6cc9ece1c92bdfda65"}
{"id": "train_074661", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e7120e84ee797756972c1057a33b67940b2cc9f3"}
{"id": "train_074662", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "4dfa9c0e8c79e33456e10f9acdb0b883ad77b281"}
{"id": "train_074663", "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\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "1d26fb2e7b7da0eaca2716b3b7745bc36b020048"}
{"id": "train_074664", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "b58017e5eb0ff2384dc3b031b397a8f8063b1570"}
{"id": "train_074665", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "d4e7f1865afa57f4f68c7803a6ac3bf0d1134e5a"}
{"id": "train_074666", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "7f65171d2af83cd3bcfa703a9001c70046922908"}
{"id": "train_074667", "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\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "4de97e22bb56ba131756845c4a60bc517f23e3c3"}
{"id": "train_074668", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "7adf01949fdb8a508e41dc7c13f8abc9abd9b6ac"}
{"id": "train_074669", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "7a32a3728e5f3bb9cfffa50d2d2e7ba7af81a03d"}
{"id": "train_074670", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "3e0073c07f144878a2384b6443f1e67a3c8f1cc4"}
{"id": "train_074671", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "f8289300c623cd2281fb8c758407aaff2ba4f920"}
{"id": "train_074672", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f140de423572970733114dc96f3b78a2bcdf7820"}
{"id": "train_074673", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "4e8609f3f1b6f7ee11418b6dca3b40835b6195bd"}
{"id": "train_074674", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "d68e325858e9c0a43a5bf52cce24f477b97df49d"}
{"id": "train_074675", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "547fcff9fc4c4cacb79498987f9d3a77bc570749"}
{"id": "train_074676", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "9db78bc4c11f42a7694f59001fc49dfca0868159"}
{"id": "train_074677", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "8dcb0f5ee57e8746ff2b9c2ea9ff7e0c930c3d32"}
{"id": "train_074678", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "d4cc4fd29f8ab7dba7b501d978529e0551836759"}
{"id": "train_074679", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "be03aa15fdca260aa1a7aeeb308dc5a18356f8f2"}
{"id": "train_074680", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "3a1bd14dfae9fad2e99b79a1b450495bf707621d"}
{"id": "train_074681", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "8a22d1b96bbd5d8b9167dcc39f2dacfc50bf66a0"}
{"id": "train_074682", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "688bd9c277ebe8c1506ca5d8c3c035c8d09fc55c"}
{"id": "train_074683", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "b6ae8f6d16af2a44cdda40ee8f12d26c6350ba6f"}
{"id": "train_074684", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "54cfaaff63b062c60c7d55126ca822a4f7c7409c"}
{"id": "train_074685", "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\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "6bf28e4314b461a5144bbb1fd4dd9b35584b2a0c"}
{"id": "train_074686", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "cac4c9009b6ab4a4e59a74c0a71b835c1cdb0bfd"}
{"id": "train_074687", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "c66b1908b068c85505619d46608c0f3f2887136b"}
{"id": "train_074688", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4ae776ad8e5b652da92f0b9ff69ecf961d4d92fd"}
{"id": "train_074689", "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\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "d9c4e35b1392fd0347e3f1f3cf8bbc9d1183f3b0"}
{"id": "train_074690", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "bdf6dba2cf9b0feb041fff864b1ddf415b9f12ab"}
{"id": "train_074691", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "1179ffb82e2fca3cce6c47eebd448f33b2a6b6bc"}
{"id": "train_074692", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "4fb3ad483c7f6566d3273cfa715a3eb926ea6ea8"}
{"id": "train_074693", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "a9756f6f9a48cc18206a2aee27f6f998ffd3b35f"}
{"id": "train_074694", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "6f0c3bc1ca64289979cd2de53254946033815e37"}
{"id": "train_074695", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "bdb8f3e3a4f763b8d85ff5a0bd16ac1f622dd390"}
{"id": "train_074696", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "cd4a8bec604f70bcc35ba06f02bbff6162e4635e"}
{"id": "train_074697", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "0d864ea8aab9b9eaec332aec0c163c655939b4c9"}
{"id": "train_074698", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "725f161412cf4a61d030fdbcab438f916c84d227"}
{"id": "train_074699", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "9c80072a33c61fd4b2f89b901cbcf416503faaf7"}
{"id": "train_074700", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "1ea562e09b7dc2b0771e72cef653bb3c97891d7a"}
{"id": "train_074701", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "3e051257227dcd8ca8855ac85af5cf82f5d8aa3a"}
{"id": "train_074702", "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\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "5aa9ab865ba54b52c68de2b63975146e24700a6d"}
{"id": "train_074703", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "2bc3592679ad57b1ed3a63ad063961ff761696b3"}
{"id": "train_074704", "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\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "e39735247d97973f8841615dd63f9dfe1c02eb5c"}
{"id": "train_074705", "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\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "237caaa61bc411f3a2e0451e666a57e7ff214ff5"}
{"id": "train_074706", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "1cbae78cb4ccd7077d97b74ee7bbfdc491f7babe"}
{"id": "train_074707", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "bca6c4dd4abf48b138b3eaf27e002d0527cdcdd2"}
{"id": "train_074708", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "ac5ed9855aa4d88d8d48c3b79c727ac48f224caf"}
{"id": "train_074709", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "b54f56556bb7eb282e4a872f6e2a9bb07a10601e"}
{"id": "train_074710", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "4cfc3a7b9ac2564315a784fd5f44d96215ecf5f9"}
{"id": "train_074711", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "8170ab7b42fc2c523345f4b98b2f7c8434a7b73e"}
{"id": "train_074712", "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\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "58086daee3833a2b85c1074c9234a452c6d021e4"}
{"id": "train_074713", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "4d9edea0bdd51f845e1d711913f641fa9b57ea2e"}
{"id": "train_074714", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["data", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "54e78e90ad14f9a9f570670910416d8ab5619a79"}
{"id": "train_074715", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "b038dcafb689fe369f36632db9e7dfa883dce6c1"}
{"id": "train_074716", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "6874136a1e4e6009d38e1b18a5e8ac935ca40c1a"}
{"id": "train_074717", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "89c9233d449e27829890261f5c5f6f0475bd12a9"}
{"id": "train_074718", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["security", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "22843ece45fc5d96225aed6730bd89f8e4bd2310"}
{"id": "train_074719", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "4ed327dc06ae3b1ab6baab3d18ccb27a90bd7494"}
{"id": "train_074720", "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\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "65f6427688a267757c107e354f776d4d37f8a56e"}
{"id": "train_074721", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "d6b2d5cdcfd44c0adcded8916c59389254b4f409"}
{"id": "train_074722", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "039684440b48c9e76933916995613f38b4d66501"}
{"id": "train_074723", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "bab37abe51abb17a20d8d4bdfcd2d5639066ecd0"}
{"id": "train_074724", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "48d60c823df17306c1ccebe12c38b6c1bc03e1ad"}
{"id": "train_074725", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e1ca50fd7242047e5d8be1918b10219237c7d2dd"}
{"id": "train_074726", "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\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "44d49c373d545450c6e0ea43878b05aaff376fc6"}
{"id": "train_074727", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "738bbb0934c65596503a183cc0ceb9a9c3b48583"}
{"id": "train_074728", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "bdd26f29d7354b633a19089f16e130ae5e3cfc64"}
{"id": "train_074729", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "780227552b2d2f617100f1c5f7ec2115a80328b0"}
{"id": "train_074730", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "23393195e24300a27a5cfd049853b06e3bf8c7bc"}
{"id": "train_074731", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "4f7f2b34665522933f7412569fdcaccc735f0d9e"}
{"id": "train_074732", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "34c88047ce976cea28b9f2eed2841e148edda472"}
{"id": "train_074733", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["governance", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e322507fa3156eceb50defb56f9ced89af4252b2"}
{"id": "train_074734", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a267dfb830372d3f3009196c08c7b954ade86295"}
{"id": "train_074735", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "6ab65e5bb97549afbf5a629d08dc25916f9d34db"}
{"id": "train_074736", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "ba615a90ca688bbd0cca2b78575607a012eddd92"}
{"id": "train_074737", "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\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "b3cf5971ac4546010cd9b0d004a90f476aef4f80"}
{"id": "train_074738", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "da7d65323ef2711cc1f031fd46ababe39cf20b57"}
{"id": "train_074739", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "f5d5fca7d9335e2867172619b4a07969ab3e7954"}
{"id": "train_074740", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "9aae8ff711150f880b3b443ace65f6f09207b0e7"}
{"id": "train_074741", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "392a7038d0a121e30cfa28bd006d221526b5de04"}
{"id": "train_074742", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f087d8d109c7398c4cfea06a091a90362a7e67a1"}
{"id": "train_074743", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "dc21e450c64a52deb5855b6d5bd7e134ada7d1fd"}
{"id": "train_074744", "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\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "14be10890e09b5e81cf79c99447878a1a0b9631b"}
{"id": "train_074745", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "3f6b7a8227b96355831be5e0c3017bb2e04d4e86"}
{"id": "train_074746", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "4c2d7e98302e1a276b7ab725b2627481d2b6f1be"}
{"id": "train_074747", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "87b6739c003ff858656581156d9268087b934f8b"}
{"id": "train_074748", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "abe6fa1b3dd64a0c24302a547b585f4863c65728"}
{"id": "train_074749", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "696a42b6a65014eae4e43cc03e908999f8454262"}
{"id": "train_074750", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "0e7321a355b62fe9949721e93ce4d4a21655bef4"}
{"id": "train_074751", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "99df876bd2808d5301ea4ed5f573b55491195f25"}
{"id": "train_074752", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["research", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "353d328011a1e17a908a7ea0bcaaeeb5fd78baa6"}
{"id": "train_074753", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "7775d1a70757ff3aaf3a6cc5609e953e50f1c916"}
{"id": "train_074754", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Rust", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "1aeca1fea86d07a2acfb1cd86b403ca0d946351a"}
{"id": "train_074755", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "d53c1afab175e8cd2822d171d74fb9e6eb4ad96f"}
{"id": "train_074756", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "251fd8e35b8df3a96c93b619f58c8dfcfaa32781"}
{"id": "train_074757", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a3a9ffcbca1e1e9778acb41329b25928faa26130"}
{"id": "train_074758", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "c04ba525844d1a45e6f50a66b193129e64aa7e0c"}
{"id": "train_074759", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "4e75d8a1027f37af352fa56204cbaeeb81a98393"}
{"id": "train_074760", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "2f68e9e07f5aef40e0efc2763a6fb321ffe8004d"}
{"id": "train_074761", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "23b7a34725d1addbf6510d0460a6d9f4fc500391"}
{"id": "train_074762", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "12fe6ac8b194c61f381f6fbc4153edd3584dafd0"}
{"id": "train_074763", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "217243d4eaebd4d94121055d99e63e20dedb9cc2"}
{"id": "train_074764", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e259007daaeff209655608019e1727b7ceca2de8"}
{"id": "train_074765", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "27aaad0cd7a8f0819b3e5842a388798c5fff5c39"}
{"id": "train_074766", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "052027b9168f83005ef5a76cfdfcec76c5f894bf"}
{"id": "train_074767", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "cff6c69d8da0f579559f01cc675038b958fb58f5"}
{"id": "train_074768", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "41da8c97baffc49d4bc7032e1a8178110a586475"}
{"id": "train_074769", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "d1a2922e625b788bb94c56d26466fd2bcf2336eb"}
{"id": "train_074770", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "7e97f658799a72a35b3c60de1a47b3de95034df7"}
{"id": "train_074771", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "5ed50501dc4eccb1180dd999b2804d8227d0bfa2"}
{"id": "train_074772", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "6f250b8ff67dcfa4f2878cd276fe250089162021"}
{"id": "train_074773", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "7c1ae225a8685b7da6afab0410f1916af24dfa15"}
{"id": "train_074774", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "6b6161f68a89e9018bd3b63951c7dad47d59b0ab"}
{"id": "train_074775", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "3822c6b4b37f3a0db20f2a3dcfbd957aaabbf379"}
{"id": "train_074776", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "8aacfbcf4a6d1aa4af7bc143abe05354037a331c"}
{"id": "train_074777", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "b510d17f6d022d37143bc7eb1c53200f4282d1ac"}
{"id": "train_074778", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "89f6ec044b894fac7663f99a2738d3ea0ca1211f"}
{"id": "train_074779", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4b75459e475b10d81c72a1ef73e06bfcbb42456d"}
{"id": "train_074780", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "cb7628838ad56508eabb38bb56cf23cbac19d9ce"}
{"id": "train_074781", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "3c2cb365191d799ee0ad7d325d8a9d7d46ff086f"}
{"id": "train_074782", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "4e561878b179666fea17bbf89aa4704abad9d114"}
{"id": "train_074783", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "8f1e62aff4aa092abb7110df3896e0a52bd6306c"}
{"id": "train_074784", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "fd7443b5cf00ccfbe7cf2189fd7abe1a1c01c3ce"}
{"id": "train_074785", "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\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "72491b00073f861742cd478154c154eef230e398"}
{"id": "train_074786", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "dde1728acbeab6f6d23e57363d2e894bbbe32cbc"}
{"id": "train_074787", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "1ad6d657ed1385aa87a323d1fb2645f2bee2a0f3"}
{"id": "train_074788", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "d18e9bb89067431922a19a4e477024330cbb9fc2"}
{"id": "train_074789", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "0b0b750b22a0bd5340bf093bf9e80cc409a9b36b"}
{"id": "train_074790", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "ead65102b08693e58ab67fb297d75f844dcd62f3"}
{"id": "train_074791", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "cb7d606e6275af479b617a87258e442d2f583841"}
{"id": "train_074792", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "6cb8b9b840709a47cca0782f87c68c8838e05bfb"}
{"id": "train_074793", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "9d6d6415d4daa4df16f5d77c6ceff735ab9302ca"}
{"id": "train_074794", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "957f0f161d42f2c28637062b52cefbff440d4c68"}
{"id": "train_074795", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "b338b44054cc8e2952b780afbca033de20b000b8"}
{"id": "train_074796", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "dff42e5f8a46f5fe83f045746eb6699db2639bef"}
{"id": "train_074797", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "fd334eb19fc427920ac50ec68c09fec9a63c59a4"}
{"id": "train_074798", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "1db480612a4866b831a912796425cc5d75036f80"}
{"id": "train_074799", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "e636044e5f1bcbfb03f4616a8641bdb8a71ed2e3"}
{"id": "train_074800", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "07278b24fac8e7f18a9a18d602dace725bb06906"}
{"id": "train_074801", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "5738cd33ebb465fb321e2483e922a338656f05b3"}
{"id": "train_074802", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "37f0d76722e455269837bdbfee9fed2635b95251"}
{"id": "train_074803", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["security", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "de13370885974319b5d397fbac6dd8448a3ec9f9"}
{"id": "train_074804", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "8621a6709b5a5ac5f48d1d12fcb3c46645e0f35b"}
{"id": "train_074805", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "6cdd45798d754268adc4cf49d5903b266fe815f4"}
{"id": "train_074806", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "01fd8a1615ad51967aefd670aa6ee5592f46b1c3"}
{"id": "train_074807", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "75c9476bc5ab4a06450a03f04598e73c28936fbe"}
{"id": "train_074808", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "0593bc73c186c92188defdb6e5ace4afb62523f2"}
{"id": "train_074809", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Python", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "51bf39ffb751a2e3a2550e94b6a00d6be38f3b17"}
{"id": "train_074810", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "0a1b8b10a94c5c0596bddbb267787956b4cd09c1"}
{"id": "train_074811", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f0cebc937c8cf796d7820bf8becd56bc646eeedb"}
{"id": "train_074812", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "6ac83ecaa6a787f054e5dc95a7c7a62b1a6f2dce"}
{"id": "train_074813", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "72cae0ef2b4a4ebf0d89a75a2f294129d0d477b8"}
{"id": "train_074814", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "7ecf727009c26db7bf37509136aca7be592fa11e"}
{"id": "train_074815", "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\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "1f5c587e3966e71aa549da3e337ee75e8ab02b79"}
{"id": "train_074816", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "d013cd06fb3e9b59b70033a09686b140759ad07e"}
{"id": "train_074817", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "c1cad9ac1d09a4f065ae4c02cf48e3e03c8cfcc5"}
{"id": "train_074818", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "06a1fba767470303887c5eae71b4c333583cb238"}
{"id": "train_074819", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "38cf7a6413b0afa3da6b742567f4b66744fea59f"}
{"id": "train_074820", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "18cc83299418af2df6a6ab265a8664b270bbcfe1"}
{"id": "train_074821", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "8ea0355905e71dcb2520ddf84f7f076ae4f6329a"}
{"id": "train_074822", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "29bb2c67b00af5482a469be40ddf8b3944975a25"}
{"id": "train_074823", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "288cb3bbe52eea6f2c2b806a4dd84532822628ee"}
{"id": "train_074824", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "baa1103dff753dedd73901a67f41d1d4e61af98a"}
{"id": "train_074825", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "01a9c279a14844dc94325ec9e2c43e890441dfa3"}
{"id": "train_074826", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "80bd1042c864d179a88f6401ea0dc8345f3d409e"}
{"id": "train_074827", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Bash", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "dd9546b0e846b22777430cc559372afba2372e66"}
{"id": "train_074828", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "b156c7e11a508ec94a4bcbc9961ad1d3ca25ecd1"}
{"id": "train_074829", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "d276faffe954c29108803d62d2ba57e6a1eda206"}
{"id": "train_074830", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "2a4f0c2eeccd41de5636749982d1fa6ad1de77cb"}
{"id": "train_074831", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "20c816091a35b7db3f693f4bc375038b4e8acafa"}
{"id": "train_074832", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["governance", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "10e55051ec0e3616dcd0928df685cf170e57fdb0"}
{"id": "train_074833", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "c3805a3fe8d639ea8b2c377fa0842e8bb0b8a4fa"}
{"id": "train_074834", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "009c86852e0325875b9ca8cd17c849e540160c7f"}
{"id": "train_074835", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "22a2214de1ed4f34cd250e7fc305cedb180d7532"}
{"id": "train_074836", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "875e3325474dff84396db7f9149544cfa93b9e70"}
{"id": "train_074837", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "90bd26758ee1b66e43eaedb5159161a36bfcb635"}
{"id": "train_074838", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "5908ec8c26b52c257898650cf5402e7154a34c05"}
{"id": "train_074839", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "5bbf0e41b1e445ddadc9c79ba2ff1757cecbe427"}
{"id": "train_074840", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "84d2f97407d444d8b42213fea999b1a4219dff03"}
{"id": "train_074841", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "4d1fce3ad3829aabfa003878207515a2a2f54b65"}
{"id": "train_074842", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "4dfa9c0e8c79e33456e10f9acdb0b883ad77b281"}
{"id": "train_074843", "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\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "c84aac87782ec8e8cfe2bcf4c8ff46b90e676178"}
{"id": "train_074844", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "93237fccaf506dfb8346c68544fd5fdfa69d48a9"}
{"id": "train_074845", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "7fd69d69ad64498409d9f70eadd928f7eefb572e"}
{"id": "train_074846", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "b5177fba56ed3e48757730f4c86b6f1e2d461c4f"}
{"id": "train_074847", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["governance", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "6b99761a761e26816caa2f42f011de80a16be14e"}
{"id": "train_074848", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "af39fe9bf50d22cd2177a3e5c0973eef18c7f41f"}
{"id": "train_074849", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "75ae19837d9bd71da2c4be66a7e11085f76136ee"}
{"id": "train_074850", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["coding", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "8cdec2ec28c28b7b567b13eec0a4fdd8312d6de2"}
{"id": "train_074851", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "ee4b99e496711962a25314f4bf4544d930c7c2b1"}
{"id": "train_074852", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "efaa4c1ab78384622755582acfe7b52bf3ce0927"}
{"id": "train_074853", "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\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "964c2c3bc25a751327c15380010a1303a7834ad5"}
{"id": "train_074854", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "45fc5e9800e3add305a60058b9e0e9b7cd5a4b27"}
{"id": "train_074855", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "e84101f2c99839ee5b4ba05e1dfd426954e2e009"}
{"id": "train_074856", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "c664fd32416037e33a135585cb20ad57b525ccae"}
{"id": "train_074857", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "250d4852dabfb80a976b82f8c02e5811d6b3aedb"}
{"id": "train_074858", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "102315d009e93378d87a7cb5a4077f8432290f95"}
{"id": "train_074859", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "75b55a5fa591b1ef59b2f6a3db6982bbd79459aa"}
{"id": "train_074860", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "d661ea578a800f7cdcfc797ccb62eae383afb48e"}
{"id": "train_074861", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "cfc4db1b593f48c405d0df59d5cf838c2356d702"}
{"id": "train_074862", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "96e850d35dd523a9a9854536406d3c260db14fc3"}
{"id": "train_074863", "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\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "479c403360873701c0f7e4c3a65b08eb96f5665e"}
{"id": "train_074864", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["security", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "fd5a9adef1acf599d4ab9f58a85250b2d3716132"}
{"id": "train_074865", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "72691c237e1049238e511f2ebac0e422f359e9bc"}
{"id": "train_074866", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 6,\n  \"confidence\": 0.6,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "7605be4586989f069759943053970fbc6212bdd2"}
{"id": "train_074867", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "54678bded36fae309b4763248cff59fc78daf49a"}
{"id": "train_074868", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "47d3da19ce702c5c99552914e2ee51ca620f77a9"}
{"id": "train_074869", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "7f6acc6c0acb1ba4ee8798d5d39bd9c9967117a4"}
{"id": "train_074870", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "e301eab1ab8d6c171d69b6c0019c06d7ca4f2774"}
{"id": "train_074871", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "6ac83ecaa6a787f054e5dc95a7c7a62b1a6f2dce"}
{"id": "train_074872", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "7e86049f9a391e9e8abc1101632e2d374603c589"}
{"id": "train_074873", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["security", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "e546d68164a14647c899ddaef0ac9028598488ae"}
{"id": "train_074874", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "ecb610f9dc67c39d2a0a7f053bb65020f8cb2036"}
{"id": "train_074875", "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\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "08d294869302c0d99bdeeb0d3ccd5399bf416fe8"}
{"id": "train_074876", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "31840c09a742121fe6a2c8ed77f8522262a366bf"}
{"id": "train_074877", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "7c13398cec1d09748bcd17faada4299581fcb87b"}
{"id": "train_074878", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "c05e6255d5e68628445752c741d5b1795c3f881d"}
{"id": "train_074879", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "eb82d99c3595c5704bbabd0e512ed5ef5fa84d51"}
{"id": "train_074880", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "707ad8d0b87ab4069049bfc1611c8646123b62af"}
{"id": "train_074881", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "e9c23641b8a773a4602c83c6597c8d03e47b1cc0"}
{"id": "train_074882", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "366c667e83e68564812c987411a7ee8157d414fe"}
{"id": "train_074883", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "a71f6d21ad14e70e5c442833c8b9ab4bbd1f6067"}
{"id": "train_074884", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "405a5d4865d00bb1c39af150846de162a17f5e9c"}
{"id": "train_074885", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "a52f95f631c5021aaf67360447ca70e602de19f7"}
{"id": "train_074886", "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\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "ad819396e0665c3e1c1be1b6e451cd741cccaa67"}
{"id": "train_074887", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "3cbadcd11b38f855b07c074b96d2fbf40baad4b2"}
{"id": "train_074888", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "1f512a98a49d700e942b31ad64427344333b9704"}
{"id": "train_074889", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "966086eaff30d93fa6d6a232bd195293cccc1dd7"}
{"id": "train_074890", "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\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "af397735b841ec543db9bf74601e857ae8f49027"}
{"id": "train_074891", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "60216b7d7633194db4caa9cd1a52cd3552e5d9d9"}
{"id": "train_074892", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "e58b4795640a7938203895d5ed5250d6add4e866"}
{"id": "train_074893", "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\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "ec442ef52052ed4701470609029f99e5dddb5a23"}
{"id": "train_074894", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "b7bce14fb4c89b7222598f0b7771a167ce595606"}
{"id": "train_074895", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "9edea02502e060fdbda2f477ffd16a1c30d0e637"}
{"id": "train_074896", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "15cf041d06c660448bb3fc6bfd090ba0be1e9a9a"}
{"id": "train_074897", "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\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "e6453511ff1d360cf2f30ee9aed61c457a00d04e"}
{"id": "train_074898", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "9c18e4083f1e5e9ed8f49bab8f7f6335227fcbab"}
{"id": "train_074899", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "727cbe452a019047a726243bb67c8f21fb33f4fc"}
{"id": "train_074900", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "2b6eae4773e1df84b2c5f554811212ddb5229123"}
{"id": "train_074901", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "794f72b6d022e8658a7b99b638a99108eba91948"}
{"id": "train_074902", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "96e850d35dd523a9a9854536406d3c260db14fc3"}
{"id": "train_074903", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "38c81a1af994daca76648cc96bbe12915c80c920"}
{"id": "train_074904", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f9465c33bc2ac9f752c23f4fd9018d900ddcf22a"}
{"id": "train_074905", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "beac60cc201acccbe5f41fd11692dcdbccf3cdcd"}
{"id": "train_074906", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "717d1f54a9e364ff5a30eeeb0226f7ea3b7c1c50"}
{"id": "train_074907", "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\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "f264bae0a39cef504b31915d9871615e76755f48"}
{"id": "train_074908", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 9,\n  \"confidence\": 0.9,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "74ad0af11c5f10148ad2773973266c17d9c64b77"}
{"id": "train_074909", "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\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "e16b8c1610c6e296067e4ddd7bfc4e6e547bb42d"}
{"id": "train_074910", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "b2f1e1987a938ad8a9d632c83d5cef9c54e0151b"}
{"id": "train_074911", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "9cac16393774a0a5a0c6fa5d71772c5668984b53"}
{"id": "train_074912", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "fed4f2b8b29375dbfc88190848caaa82342b6852"}
{"id": "train_074913", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "988e5b61b382bac6126428175ace83377d02c876"}
{"id": "train_074914", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "d7e02ccda21533a39e6b05d932de08220a96c123"}
{"id": "train_074915", "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\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "ed934caab1ccfc5d7cbe8b9ea6dbabceb6d04866"}
{"id": "train_074916", "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\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n    cost = 0\n    for _ in range(budget):\n        p = plan(issue)\n        patch = edit(issue, p)\n        ok, _ = test(patch)\n        cost += 1\n        if ok:\n            return patch, cost\n    raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "7b85afb6f163821545a158023bf8f71f96ad06ba"}
{"id": "train_074917", "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\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "184d5f486d243a6d163851c353d08e963a9a25bd"}
{"id": "train_074918", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "8c84edbdbda008d1623dccba98a505d75db23629"}
{"id": "train_074919", "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\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "a59df88bedd438e17359c75f157799522197669c"}
{"id": "train_074920", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "3a37a172a720639ba0536dc133585e7258fa4734"}
{"id": "train_074921", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "d035dee82db109fc575594ed1e7424c6636dbd5b"}
{"id": "train_074922", "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\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "1bb8d723c0df1c9ee401b31571c72725b2ea72cb"}
{"id": "train_074923", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "70ad2e10ca304988d446bda2cfaafa12f9536ffa"}
{"id": "train_074924", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["governance", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "6e74c838e27c32dd5e491115a2b7fc8a5c6b4bcd"}
{"id": "train_074925", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "e78376fcb9271790c240616f99e9b7593b52978f"}
{"id": "train_074926", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "C#", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "56e27822858b0ffa3b9a2ae4f867b15cfe2a4736"}
{"id": "train_074927", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f8298d41c50401d373cc476adf9f1edfcf28353f"}
{"id": "train_074928", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "a4dbc7384d360eb86611d5787f6b133a87279497"}
{"id": "train_074929", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "f2538e47683711ca313a35ca9202848a09454b61"}
{"id": "train_074930", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "a24a0079d7c7b48da351c67ee247beda5ad56a44"}
{"id": "train_074931", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "7f587224784918a0ebc41b319feef146227dcf23"}
{"id": "train_074932", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "538f2a57dfe1c5a5d73b915279ae472d40abe80a"}
{"id": "train_074933", "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\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "ff6ead12fdc96534d2c6c91c543d73b6e6fc5659"}
{"id": "train_074934", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a5f8df7329d4ff36e7f99c79684ea3ecfc6400cb"}
{"id": "train_074935", "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\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "9a63a0ae61b9966f06740d23f08a8dcca06e1ba8"}
{"id": "train_074936", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "89dfddb695d097fffaeabf3b6c8ec7760d4ea7c9"}
{"id": "train_074937", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a2fbbce6a4e7413d86337f57aec831f1c5bc80b1"}
{"id": "train_074938", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "1d6a0357c276e1fb2e693675c73a1e204cbf2d4c"}
{"id": "train_074939", "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\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "178548142c851fc6febb65768198d94b6d69fc87"}
{"id": "train_074940", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "8526d955e0f77a9d8a64108537f0e0e33fdb80c1"}
{"id": "train_074941", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "5a6a87bc11367d4dc7c8120c5eeba80889c7b73f"}
{"id": "train_074942", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["agent", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "688307756e278c3e7b782115d4796db11b68705f"}
{"id": "train_074943", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "9dfd330b5595c10aeb933e826a1a0c753a29f593"}
{"id": "train_074944", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "8a722794f8061e0011b2b95cea02eb58f308bc74"}
{"id": "train_074945", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "2a7d7e2e097419834d7697c4f61049a24feaeeba"}
{"id": "train_074946", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "SQL", "moe_experts": ["coding", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "a6f2f0edb39c7de16dde3afff85b8f9d7c1f4af9"}
{"id": "train_074947", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "11ad04fde999b7b33e31f0116d4819ec31533068"}
{"id": "train_074948", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "3f7d7a9534f66b9d0eb66a6220614fadf0467dec"}
{"id": "train_074949", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 8,\n  \"confidence\": 0.8,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["coding", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "d2539ac79a2678a0c6cad6c5a42a0c5f22a60f1c"}
{"id": "train_074950", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "11149c4b1237fe8581234fd99a5d68b97088ffde"}
{"id": "train_074951", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "3ec9c5b454ab970233ffc0fb64ddb5d3f20f8e04"}
{"id": "train_074952", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "29cbd4deae69394b34eea2d25c010256aad71d75"}
{"id": "train_074953", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "42a16cdc3ef0774fcc7b257098bae8ab36bf1f80"}
{"id": "train_074954", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Java", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "73c58e775427747ae08650d459378182dad7ab0d"}
{"id": "train_074955", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "5200801a6b22b8a4ba7813b1b65374fd229886d2"}
{"id": "train_074956", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "903618e154625e050c6f30830cac62205f5ddbcd"}
{"id": "train_074957", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "e590c630a2f63e635d1377e0febf0b9b91d1f554"}
{"id": "train_074958", "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\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "f9eea076940cb3affea538aed7ed9d95f4aefbe2"}
{"id": "train_074959", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "52ee8812434424a43de894bbb892aadf5fea7e02"}
{"id": "train_074960", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f12811ce14caab05c315e36fa8b4a0c91dcf24ca"}
{"id": "train_074961", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "aa8107936b2f93c4e305c95bf96e7afefce10882"}
{"id": "train_074962", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "872f3aac3c9f1fded40aefb9016e03a7e48076cf"}
{"id": "train_074963", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "05a2950bae1ee0126a35f406d2914325e60a5727"}
{"id": "train_074964", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "ff3d87ef688ea2e616ff1300bd6e332f55e7faca"}
{"id": "train_074965", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "eba3182ee98952702b2e24a954e86b32f9359cc3"}
{"id": "train_074966", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e4c10beeffb6ac09a34873b45c221e897f96a46f"}
{"id": "train_074967", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "d40830cc3ab34d9256c5792164e20ac71bcfc0f8"}
{"id": "train_074968", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "955c0969cb623896eca5917ed31015d8663c1877"}
{"id": "train_074969", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "a914a3a3b8ea306b53f7a1bd9f6f8a56f4eeb6b0"}
{"id": "train_074970", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "dec5fbf2d07f86149230108d072beed0ed6367ec"}
{"id": "train_074971", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "80853098eef4c055e13924d08f1f07fb32d47872"}
{"id": "train_074972", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "756a93f0aa19dd5fb48fe61d5bd1574dbec51dc2"}
{"id": "train_074973", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 10,\n  \"confidence\": 1.0,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "72440cd79362431077b00bd5a7f8290bbdb54f96"}
{"id": "train_074974", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "c0b5d13086d4a7faa06ce629efe7396d1c5b9610"}
{"id": "train_074975", "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\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+     raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e823d97d25c01f1980221e1a6fc553bbfa30f86a"}
{"id": "train_074976", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "5d7eab3caa0d3c3d53fff8bacc5a3760ab58a036"}
{"id": "train_074977", "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\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "7434cd007340c5c18fa6e18e63c42e26f7864bf6"}
{"id": "train_074978", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "a04c21fdb7a770cdeb5c23bc1d25810519647b1b"}
{"id": "train_074979", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "5bd361a0ed5cf1036bc3598be27deec3bcc07332"}
{"id": "train_074980", "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\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "cce79a4443c15f6f38b8370c3d08fc98bef6e46b"}
{"id": "train_074981", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "dcd895150cda878fdfc6b63d43a84fccfe35eb1b"}
{"id": "train_074982", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a78a3a30f993f7160e9f7961e2382a74bd5ce9f4"}
{"id": "train_074983", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "d83e7f88b548409404a6d419cb14a1c411f82dee"}
{"id": "train_074984", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "47589ea8eb329f5bc732eb7bfbaae0dce7fe7570"}
{"id": "train_074985", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "e4abd09338b4118de2c3c4f495bba8736ef22484"}
{"id": "train_074986", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "016ab9ffafecbab94afa1f2f12049015e179b480"}
{"id": "train_074987", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "1ef7f3e3cba4846fcfd9eee35f213b3d07dcdddc"}
{"id": "train_074988", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "188a396f3d4be5e08d7b4b68263ab139bf6068e6"}
{"id": "train_074989", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["security", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "ee1b0fe37730cc0d34d9475b8472eb5305464b72"}
{"id": "train_074990", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "0601d44909b4d7e6e59c6d99da119cf30255a6f7"}
{"id": "train_074991", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "5930b51a4c910164becdc40eb9b639c98ca386c1"}
{"id": "train_074992", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "954f3976e0babb2d989f9c6d95b3bdec442bdccf"}
{"id": "train_074993", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "9c12f9b0d8d00dd24913f09023e636a6e768d3e4"}
{"id": "train_074994", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n  \"score\": 7,\n  \"confidence\": 0.7,\n  \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "de4561d51679353c3141c977b100ecc5f6daa113"}
{"id": "train_074995", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "49d5f095775a18a39ac69e25258887627dd11941"}
{"id": "train_074996", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n  {\n    \"tool\": \"repo_checkout\",\n    \"args\": {\n      \"commit\": \"abc123\"\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"unit\"\n    }\n  },\n  {\n    \"tool\": \"apply_patch\",\n    \"args\": {\n      \"files\": 1\n    }\n  },\n  {\n    \"tool\": \"run_tests\",\n    \"args\": {\n      \"suite\": \"full\"\n    }\n  }\n]", "metadata": {"language": "Go", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "525e707f4c88649f0dd081b226343deae5e0c250"}
{"id": "train_074997", "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\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "0fe4054cac125ec34cd42806d6a99964d70d7fad"}
{"id": "train_074998", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "409823b28396a302994eb18626fd68f105c08204"}
{"id": "train_074999", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "75ce1f13dacf3d1f7314cbbe4f2b588e0346af66"}