id
stringlengths
11
11
created
timestamp[s]date
2026-01-01 00:00:00
2026-01-01 00:00:00
topic
stringclasses
14 values
task_type
stringclasses
10 values
difficulty
stringclasses
3 values
instruction
stringlengths
189
248
input
stringclasses
1 value
output
stringclasses
9 values
reasoning_steps
listlengths
0
5
metadata
dict
hash
stringlengths
40
40
train_20200
2026-01-01T00:00:00
Tool calling, sandboxes, and CI integration
eval
expert
Task: eval Topic: Tool calling, sandboxes, and CI integration Difficulty: expert Target language: TypeScript Context: High-traffic service with latency SLOs. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Eval: pass@k, time-to-green, regressions, diff size
[]
{ "target_language": "TypeScript", "developer_needs": [ "tooling", "reproducibility", "documentation", "cost_latency_tradeoffs" ], "moe_experts": [ "agentic_systems_expert", "evaluation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
0571532701c083f0e60a8e3d5d3797dcc7a354bd
train_20201
2026-01-01T00:00:00
Agentic coding systems (plan→edit→test→reflect)
agent_loop
expert
Task: agent_loop Topic: Agentic coding systems (plan→edit→test→reflect) Difficulty: expert Target language: Go Context: Large monorepo with flaky tests and strict CI. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Loop: Plan → Edit → Test → Reflect → Human gate
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "Go", "developer_needs": [ "tests_are_truth", "repo_scale_reasoning", "tooling", "ci_integration" ], "moe_experts": [ "governance_expert", "security_expert" ], "governance": { "audit_required": true, "tests_required": true } }
b608634be04a17337913f4f80f5aa505b9d0432a
train_20202
2026-01-01T00:00:00
Reasoning-first coding models and tunable deliberation
agent_loop
intermediate
Task: agent_loop Topic: Reasoning-first coding models and tunable deliberation Difficulty: intermediate Target language: C# Context: Large monorepo with flaky tests and strict CI. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Loop: Plan → Edit → Test → Reflect → Human gate
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "C#", "developer_needs": [ "tests_are_truth", "tooling", "governance", "auditability" ], "moe_experts": [ "data_curation_expert", "evaluation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
87d3e062601106c0289416430f670ba35bac1212
train_20203
2026-01-01T00:00:00
Extended context and repo-scale understanding
patch_diff
advanced
Task: patch_diff Topic: Extended context and repo-scale understanding Difficulty: advanced Target language: Go Context: Large monorepo with flaky tests and strict CI. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Patch (diff-style): ```diff - if x == 0: - return 1/x + if x == 0: + raise ValueError('division by zero') ``` Acceptance: - Tests pass - No new regressions
[]
{ "target_language": "Go", "developer_needs": [ "tooling", "repo_scale_reasoning", "documentation", "reproducibility" ], "moe_experts": [ "performance_expert", "agentic_systems_expert" ], "governance": { "audit_required": true, "tests_required": true } }
5d571774cbf6c9dd50318a69ce74e2cfb3eda4c8
train_20204
2026-01-01T00:00:00
Agentic coding systems (plan→edit→test→reflect)
code
expert
Task: code Topic: Agentic coding systems (plan→edit→test→reflect) Difficulty: expert Target language: C# Context: Offline/local deployment with limited compute. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Design with risks, metrics, acceptance criteria
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "C#", "developer_needs": [ "repo_scale_reasoning", "security_gates", "governance", "ci_integration" ], "moe_experts": [ "data_curation_expert", "security_expert" ], "governance": { "audit_required": true, "tests_required": true } }
3a8c6c0f6adf28ac248922d79d0e298656f156df
train_20205
2026-01-01T00:00:00
Model merging, distillation, and continued pretraining
review
advanced
Task: review Topic: Model merging, distillation, and continued pretraining Difficulty: advanced Target language: Go Context: Large monorepo with flaky tests and strict CI. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Review: correctness, security, performance, governance
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "Go", "developer_needs": [ "tooling", "tests_are_truth", "repo_scale_reasoning", "governance" ], "moe_experts": [ "security_expert", "evaluation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
aa7f4e6cef1e5bdcffb87e7f42760b06d17cf5fa
train_20206
2026-01-01T00:00:00
Mixture-of-Experts (MoE) for code
data_pipeline
expert
Task: data_pipeline Topic: Mixture-of-Experts (MoE) for code Difficulty: expert Target language: Go Context: Research team validating claims against real repos. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Pipeline: Ingest → Normalize → Filter → Dedupe → Quality → Mix → Audit
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "Go", "developer_needs": [ "reproducibility", "evaluation_metrics", "documentation", "governance" ], "moe_experts": [ "coding_expert", "performance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
749ed6708588f0537cdb48d9878d9e6d8f837b22
train_20207
2026-01-01T00:00:00
Model merging, distillation, and continued pretraining
compare
expert
Task: compare Topic: Model merging, distillation, and continued pretraining Difficulty: expert Target language: Python Context: Offline/local deployment with limited compute. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Compare: capability, cost, latency, reliability
[]
{ "target_language": "Python", "developer_needs": [ "evaluation_metrics", "documentation", "tests_are_truth", "repo_scale_reasoning" ], "moe_experts": [ "security_expert", "agentic_systems_expert" ], "governance": { "audit_required": true, "tests_required": true } }
1399d2be5e46137f37d2c819662d7775b0b17759
train_20208
2026-01-01T00:00:00
Reasoning-first coding models and tunable deliberation
code
expert
Task: code Topic: Reasoning-first coding models and tunable deliberation Difficulty: expert Target language: Go Context: Offline/local deployment with limited compute. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Design with risks, metrics, acceptance criteria
[]
{ "target_language": "Go", "developer_needs": [ "security_gates", "tests_are_truth", "documentation", "reproducibility" ], "moe_experts": [ "governance_expert", "data_curation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
5802a0f98c087a0c20648ae2ed809b286c4ee5f1
train_20209
2026-01-01T00:00:00
Model merging, distillation, and continued pretraining
explain
expert
Task: explain Topic: Model merging, distillation, and continued pretraining Difficulty: expert Target language: Python Context: Research team validating claims against real repos. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Design with risks, metrics, acceptance criteria
[]
{ "target_language": "Python", "developer_needs": [ "tooling", "auditability", "governance", "repo_scale_reasoning" ], "moe_experts": [ "data_curation_expert", "performance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
4ec7012b56c6ae9d158b5d6338fb2f2e6a6769f1
train_20210
2026-01-01T00:00:00
Extended context and repo-scale understanding
data_pipeline
expert
Task: data_pipeline Topic: Extended context and repo-scale understanding Difficulty: expert Target language: JavaScript Context: Regulated environment requiring audit trails. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Pipeline: Ingest → Normalize → Filter → Dedupe → Quality → Mix → Audit
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "JavaScript", "developer_needs": [ "evaluation_metrics", "reproducibility", "documentation", "security_gates" ], "moe_experts": [ "agentic_systems_expert", "coding_expert" ], "governance": { "audit_required": true, "tests_required": true } }
41508e594d4a29037cb9af4b56dae65712e7bb5c
train_20211
2026-01-01T00:00:00
Dataset curation pipelines (filter, dedupe, quality)
failure_analysis
advanced
Task: failure_analysis Topic: Dataset curation pipelines (filter, dedupe, quality) Difficulty: advanced Target language: C# Context: Regulated environment requiring audit trails. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Failure: - Initial patch broke edge case Reflection: - Missing zero-input guard Correction: - Add explicit validation + test
[]
{ "target_language": "C#", "developer_needs": [ "documentation", "repo_scale_reasoning", "ci_integration", "security_gates" ], "moe_experts": [ "governance_expert", "evaluation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
34c5122b6ea3ebf4d89d152ec30bb3b0ba1b3602
train_20212
2026-01-01T00:00:00
Latency, cost, and reliability optimization
code
intermediate
Task: code Topic: Latency, cost, and reliability optimization Difficulty: intermediate Target language: C# Context: Regulated environment requiring audit trails. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Design with risks, metrics, acceptance criteria
[]
{ "target_language": "C#", "developer_needs": [ "tooling", "security_gates", "cost_latency_tradeoffs", "repo_scale_reasoning" ], "moe_experts": [ "performance_expert", "security_expert" ], "governance": { "audit_required": true, "tests_required": true } }
b6c8036cc389404d23adfa7dbd8edda1b81f6884
train_20213
2026-01-01T00:00:00
Tool calling, sandboxes, and CI integration
review
expert
Task: review Topic: Tool calling, sandboxes, and CI integration Difficulty: expert Target language: Go Context: High-traffic service with latency SLOs. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Review: correctness, security, performance, governance
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "Go", "developer_needs": [ "documentation", "tooling", "ci_integration", "reproducibility" ], "moe_experts": [ "data_curation_expert", "agentic_systems_expert" ], "governance": { "audit_required": true, "tests_required": true } }
6f4d4aa18e44f85a31930d2f3ad5cd75e0dcd604
train_20214
2026-01-01T00:00:00
Agentic coding systems (plan→edit→test→reflect)
agent_loop
expert
Task: agent_loop Topic: Agentic coding systems (plan→edit→test→reflect) Difficulty: expert Target language: Go Context: Regulated environment requiring audit trails. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Loop: Plan → Edit → Test → Reflect → Human gate
[]
{ "target_language": "Go", "developer_needs": [ "governance", "ci_integration", "documentation", "tooling" ], "moe_experts": [ "governance_expert", "evaluation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
46570a4b7ae39521521f933b35dd7d7e65428ad1
train_20215
2026-01-01T00:00:00
Multimodal dev workflows (docs, diagrams, traces)
agent_loop
advanced
Task: agent_loop Topic: Multimodal dev workflows (docs, diagrams, traces) Difficulty: advanced Target language: C# Context: Research team validating claims against real repos. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Loop: Plan → Edit → Test → Reflect → Human gate
[]
{ "target_language": "C#", "developer_needs": [ "repo_scale_reasoning", "tests_are_truth", "ci_integration", "security_gates" ], "moe_experts": [ "evaluation_expert", "data_curation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
41eed858eee6a9f8726302da4d65dc56803b3b63
train_20216
2026-01-01T00:00:00
Dataset curation pipelines (filter, dedupe, quality)
data_pipeline
intermediate
Task: data_pipeline Topic: Dataset curation pipelines (filter, dedupe, quality) Difficulty: intermediate Target language: SQL Context: Research team validating claims against real repos. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Pipeline: Ingest → Normalize → Filter → Dedupe → Quality → Mix → Audit
[]
{ "target_language": "SQL", "developer_needs": [ "auditability", "tooling", "cost_latency_tradeoffs", "security_gates" ], "moe_experts": [ "coding_expert", "data_curation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
da9c5fce130d8c35ef4ecc62c92189ed7cafe3f4
train_20217
2026-01-01T00:00:00
Model merging, distillation, and continued pretraining
design
intermediate
Task: design Topic: Model merging, distillation, and continued pretraining Difficulty: intermediate Target language: Rust Context: High-traffic service with latency SLOs. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Design with risks, metrics, acceptance criteria
[]
{ "target_language": "Rust", "developer_needs": [ "evaluation_metrics", "governance", "documentation", "tooling" ], "moe_experts": [ "evaluation_expert", "coding_expert" ], "governance": { "audit_required": true, "tests_required": true } }
85105b0ed13ce03c811aa18c38a41708c3c0450a
train_20218
2026-01-01T00:00:00
Governance, provenance, and licensing for code data
agent_loop
expert
Task: agent_loop Topic: Governance, provenance, and licensing for code data Difficulty: expert Target language: Java Context: Large monorepo with flaky tests and strict CI. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Loop: Plan → Edit → Test → Reflect → Human gate
[]
{ "target_language": "Java", "developer_needs": [ "auditability", "security_gates", "tooling", "tests_are_truth" ], "moe_experts": [ "performance_expert", "coding_expert" ], "governance": { "audit_required": true, "tests_required": true } }
0fd0c78d9ec0d4546c1c2de6568df2f6ac66452c
train_20219
2026-01-01T00:00:00
SWE-bench style real-repo evaluation
data_pipeline
advanced
Task: data_pipeline Topic: SWE-bench style real-repo evaluation Difficulty: advanced Target language: Rust Context: Research team validating claims against real repos. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Pipeline: Ingest → Normalize → Filter → Dedupe → Quality → Mix → Audit
[]
{ "target_language": "Rust", "developer_needs": [ "tooling", "repo_scale_reasoning", "ci_integration", "evaluation_metrics" ], "moe_experts": [ "performance_expert", "coding_expert" ], "governance": { "audit_required": true, "tests_required": true } }
53d5c25a770e210f2aba897e9b15ffd78e9a9544
train_20220
2026-01-01T00:00:00
Self-improving agents and feedback loops
patch_diff
intermediate
Task: patch_diff Topic: Self-improving agents and feedback loops Difficulty: intermediate Target language: Python Context: Research team validating claims against real repos. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Patch (diff-style): ```diff - if x == 0: - return 1/x + if x == 0: + raise ValueError('division by zero') ``` Acceptance: - Tests pass - No new regressions
[]
{ "target_language": "Python", "developer_needs": [ "governance", "ci_integration", "tests_are_truth", "documentation" ], "moe_experts": [ "agentic_systems_expert", "governance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
6e3017df39fba547d97bd3c6668f954455f62d17
train_20221
2026-01-01T00:00:00
Reasoning-first coding models and tunable deliberation
code
advanced
Task: code Topic: Reasoning-first coding models and tunable deliberation Difficulty: advanced Target language: Java Context: High-traffic service with latency SLOs. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Design with risks, metrics, acceptance criteria
[]
{ "target_language": "Java", "developer_needs": [ "ci_integration", "tests_are_truth", "documentation", "repo_scale_reasoning" ], "moe_experts": [ "security_expert", "agentic_systems_expert" ], "governance": { "audit_required": true, "tests_required": true } }
bde8b487ae3a24009f6a997fcecbc52d3240ec16
train_20222
2026-01-01T00:00:00
Multimodal dev workflows (docs, diagrams, traces)
eval
intermediate
Task: eval Topic: Multimodal dev workflows (docs, diagrams, traces) Difficulty: intermediate Target language: Go Context: Large monorepo with flaky tests and strict CI. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Eval: pass@k, time-to-green, regressions, diff size
[]
{ "target_language": "Go", "developer_needs": [ "reproducibility", "cost_latency_tradeoffs", "documentation", "auditability" ], "moe_experts": [ "agentic_systems_expert", "evaluation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
8573db12577bc47d449d78a0044464068155ca5b
train_20223
2026-01-01T00:00:00
Mixture-of-Experts (MoE) for code
patch_diff
advanced
Task: patch_diff Topic: Mixture-of-Experts (MoE) for code Difficulty: advanced Target language: Java Context: Offline/local deployment with limited compute. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Patch (diff-style): ```diff - if x == 0: - return 1/x + if x == 0: + raise ValueError('division by zero') ``` Acceptance: - Tests pass - No new regressions
[]
{ "target_language": "Java", "developer_needs": [ "cost_latency_tradeoffs", "reproducibility", "documentation", "security_gates" ], "moe_experts": [ "governance_expert", "evaluation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
7701138d5a33d5a498fa9b90c0bc0e8fd2978a66
train_20224
2026-01-01T00:00:00
Agentic coding systems (plan→edit→test→reflect)
agent_loop
intermediate
Task: agent_loop Topic: Agentic coding systems (plan→edit→test→reflect) Difficulty: intermediate Target language: Java Context: Offline/local deployment with limited compute. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Loop: Plan → Edit → Test → Reflect → Human gate
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "Java", "developer_needs": [ "auditability", "repo_scale_reasoning", "reproducibility", "evaluation_metrics" ], "moe_experts": [ "coding_expert", "evaluation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
a09caa2f37b800df3c1fc740c001bac9fbbae498
train_20225
2026-01-01T00:00:00
Mixture-of-Experts (MoE) for code
explain
advanced
Task: explain Topic: Mixture-of-Experts (MoE) for code Difficulty: advanced Target language: Rust Context: Large monorepo with flaky tests and strict CI. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Design with risks, metrics, acceptance criteria
[]
{ "target_language": "Rust", "developer_needs": [ "security_gates", "tooling", "evaluation_metrics", "cost_latency_tradeoffs" ], "moe_experts": [ "evaluation_expert", "coding_expert" ], "governance": { "audit_required": true, "tests_required": true } }
fe5dfd6978f507ec70deb07a0df401ffdb73a9dc
train_20226
2026-01-01T00:00:00
Latency, cost, and reliability optimization
eval
expert
Task: eval Topic: Latency, cost, and reliability optimization Difficulty: expert Target language: Java Context: Research team validating claims against real repos. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Eval: pass@k, time-to-green, regressions, diff size
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "Java", "developer_needs": [ "ci_integration", "tooling", "repo_scale_reasoning", "reproducibility" ], "moe_experts": [ "governance_expert", "evaluation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
ede7fc5bb114ffb3a510e76ae03dae122812d496
train_20227
2026-01-01T00:00:00
Dataset curation pipelines (filter, dedupe, quality)
compare
intermediate
Task: compare Topic: Dataset curation pipelines (filter, dedupe, quality) Difficulty: intermediate Target language: Python Context: Research team validating claims against real repos. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Compare: capability, cost, latency, reliability
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "Python", "developer_needs": [ "evaluation_metrics", "documentation", "tests_are_truth", "auditability" ], "moe_experts": [ "performance_expert", "coding_expert" ], "governance": { "audit_required": true, "tests_required": true } }
3f1950202571672404226cb4b67328328074e271
train_20228
2026-01-01T00:00:00
Latency, cost, and reliability optimization
compare
expert
Task: compare Topic: Latency, cost, and reliability optimization Difficulty: expert Target language: Java Context: Large monorepo with flaky tests and strict CI. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Compare: capability, cost, latency, reliability
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "Java", "developer_needs": [ "repo_scale_reasoning", "tooling", "evaluation_metrics", "ci_integration" ], "moe_experts": [ "coding_expert", "security_expert" ], "governance": { "audit_required": true, "tests_required": true } }
00b9a6999679bdcd3c0a811b2eb4a04a203069bf
train_20229
2026-01-01T00:00:00
Dataset curation pipelines (filter, dedupe, quality)
failure_analysis
expert
Task: failure_analysis Topic: Dataset curation pipelines (filter, dedupe, quality) Difficulty: expert Target language: Go Context: Large monorepo with flaky tests and strict CI. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Failure: - Initial patch broke edge case Reflection: - Missing zero-input guard Correction: - Add explicit validation + test
[]
{ "target_language": "Go", "developer_needs": [ "documentation", "repo_scale_reasoning", "ci_integration", "security_gates" ], "moe_experts": [ "governance_expert", "data_curation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
85ff536d3f5b2bbcfce3796b57596b5c50684016
train_20230
2026-01-01T00:00:00
Self-improving agents and feedback loops
explain
intermediate
Task: explain Topic: Self-improving agents and feedback loops Difficulty: intermediate Target language: Python Context: Regulated environment requiring audit trails. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Design with risks, metrics, acceptance criteria
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "Python", "developer_needs": [ "tests_are_truth", "auditability", "ci_integration", "tooling" ], "moe_experts": [ "agentic_systems_expert", "performance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
b7df3b409f5e8e72ed85f8b4434195d64cc602c2
train_20231
2026-01-01T00:00:00
Code-specialized model families and sizing tradeoffs
data_pipeline
advanced
Task: data_pipeline Topic: Code-specialized model families and sizing tradeoffs Difficulty: advanced Target language: Python Context: Research team validating claims against real repos. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Pipeline: Ingest → Normalize → Filter → Dedupe → Quality → Mix → Audit
[]
{ "target_language": "Python", "developer_needs": [ "ci_integration", "evaluation_metrics", "reproducibility", "documentation" ], "moe_experts": [ "governance_expert", "performance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
9ad07bc2fc525048b29b6e420ffb79f2f72d706f
train_20232
2026-01-01T00:00:00
Latency, cost, and reliability optimization
failure_analysis
advanced
Task: failure_analysis Topic: Latency, cost, and reliability optimization Difficulty: advanced Target language: Python Context: Research team validating claims against real repos. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Failure: - Initial patch broke edge case Reflection: - Missing zero-input guard Correction: - Add explicit validation + test
[]
{ "target_language": "Python", "developer_needs": [ "evaluation_metrics", "governance", "tooling", "security_gates" ], "moe_experts": [ "governance_expert", "evaluation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
dba62d81b4cd81d8977fc81423b091aa68c928d3
train_20233
2026-01-01T00:00:00
SWE-bench style real-repo evaluation
agent_loop
expert
Task: agent_loop Topic: SWE-bench style real-repo evaluation Difficulty: expert Target language: C# Context: Research team validating claims against real repos. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Loop: Plan → Edit → Test → Reflect → Human gate
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "C#", "developer_needs": [ "ci_integration", "evaluation_metrics", "tests_are_truth", "reproducibility" ], "moe_experts": [ "governance_expert", "data_curation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
330485acdfbce80307c363790e38ddfba30aa4e0
train_20234
2026-01-01T00:00:00
Self-improving agents and feedback loops
data_pipeline
advanced
Task: data_pipeline Topic: Self-improving agents and feedback loops Difficulty: advanced Target language: JavaScript Context: Offline/local deployment with limited compute. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Pipeline: Ingest → Normalize → Filter → Dedupe → Quality → Mix → Audit
[]
{ "target_language": "JavaScript", "developer_needs": [ "cost_latency_tradeoffs", "ci_integration", "documentation", "tooling" ], "moe_experts": [ "evaluation_expert", "security_expert" ], "governance": { "audit_required": true, "tests_required": true } }
81591686af2d3ca22e0b384b56b27568058401d7
train_20235
2026-01-01T00:00:00
Secure code generation and policy gates
eval
intermediate
Task: eval Topic: Secure code generation and policy gates Difficulty: intermediate Target language: TypeScript Context: Regulated environment requiring audit trails. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Eval: pass@k, time-to-green, regressions, diff size
[]
{ "target_language": "TypeScript", "developer_needs": [ "reproducibility", "security_gates", "tests_are_truth", "cost_latency_tradeoffs" ], "moe_experts": [ "data_curation_expert", "governance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
5a0f12828f5bd646d47b9a400b12f3bfbb73a98f
train_20236
2026-01-01T00:00:00
Extended context and repo-scale understanding
design
intermediate
Task: design Topic: Extended context and repo-scale understanding Difficulty: intermediate Target language: Java Context: Research team validating claims against real repos. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Design with risks, metrics, acceptance criteria
[]
{ "target_language": "Java", "developer_needs": [ "tooling", "ci_integration", "documentation", "auditability" ], "moe_experts": [ "coding_expert", "data_curation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
af497c93dc082ec822c8a4aa3341a2ca6d3fce10
train_20237
2026-01-01T00:00:00
Code-specialized model families and sizing tradeoffs
agent_loop
intermediate
Task: agent_loop Topic: Code-specialized model families and sizing tradeoffs Difficulty: intermediate Target language: Rust Context: Regulated environment requiring audit trails. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Loop: Plan → Edit → Test → Reflect → Human gate
[]
{ "target_language": "Rust", "developer_needs": [ "governance", "evaluation_metrics", "tests_are_truth", "cost_latency_tradeoffs" ], "moe_experts": [ "performance_expert", "evaluation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
7591b0796f3c4eb18841ba171b3250101c20cede
train_20238
2026-01-01T00:00:00
Reasoning-first coding models and tunable deliberation
design
expert
Task: design Topic: Reasoning-first coding models and tunable deliberation Difficulty: expert Target language: SQL Context: High-traffic service with latency SLOs. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Design with risks, metrics, acceptance criteria
[]
{ "target_language": "SQL", "developer_needs": [ "governance", "evaluation_metrics", "tooling", "repo_scale_reasoning" ], "moe_experts": [ "security_expert", "data_curation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
26d903020975cf29a13d1cf4a85e90da1ed159c9
train_20239
2026-01-01T00:00:00
Tool calling, sandboxes, and CI integration
patch_diff
intermediate
Task: patch_diff Topic: Tool calling, sandboxes, and CI integration Difficulty: intermediate Target language: JavaScript Context: Offline/local deployment with limited compute. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Patch (diff-style): ```diff - if x == 0: - return 1/x + if x == 0: + raise ValueError('division by zero') ``` Acceptance: - Tests pass - No new regressions
[]
{ "target_language": "JavaScript", "developer_needs": [ "reproducibility", "governance", "ci_integration", "cost_latency_tradeoffs" ], "moe_experts": [ "coding_expert", "data_curation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
4c75496836e82413bbb0edae9888a6f99c499ba3
train_20240
2026-01-01T00:00:00
Secure code generation and policy gates
data_pipeline
advanced
Task: data_pipeline Topic: Secure code generation and policy gates Difficulty: advanced Target language: JavaScript Context: Regulated environment requiring audit trails. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Pipeline: Ingest → Normalize → Filter → Dedupe → Quality → Mix → Audit
[]
{ "target_language": "JavaScript", "developer_needs": [ "governance", "security_gates", "repo_scale_reasoning", "tooling" ], "moe_experts": [ "security_expert", "governance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
11056e32456dbf44d1514d9a22e9cab18ac2ee49
train_20241
2026-01-01T00:00:00
SWE-bench style real-repo evaluation
data_pipeline
expert
Task: data_pipeline Topic: SWE-bench style real-repo evaluation Difficulty: expert Target language: SQL Context: Research team validating claims against real repos. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Pipeline: Ingest → Normalize → Filter → Dedupe → Quality → Mix → Audit
[]
{ "target_language": "SQL", "developer_needs": [ "auditability", "ci_integration", "cost_latency_tradeoffs", "documentation" ], "moe_experts": [ "coding_expert", "security_expert" ], "governance": { "audit_required": true, "tests_required": true } }
41bb834c2ae04164c073261a1e68fd9b406b62c8
train_20242
2026-01-01T00:00:00
Latency, cost, and reliability optimization
failure_analysis
advanced
Task: failure_analysis Topic: Latency, cost, and reliability optimization Difficulty: advanced Target language: Rust Context: Regulated environment requiring audit trails. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Failure: - Initial patch broke edge case Reflection: - Missing zero-input guard Correction: - Add explicit validation + test
[]
{ "target_language": "Rust", "developer_needs": [ "governance", "tooling", "cost_latency_tradeoffs", "auditability" ], "moe_experts": [ "security_expert", "governance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
259f31a334945e8207be0440dc91b8d56356654c
train_20243
2026-01-01T00:00:00
Model merging, distillation, and continued pretraining
explain
intermediate
Task: explain Topic: Model merging, distillation, and continued pretraining Difficulty: intermediate Target language: JavaScript Context: Research team validating claims against real repos. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Design with risks, metrics, acceptance criteria
[]
{ "target_language": "JavaScript", "developer_needs": [ "repo_scale_reasoning", "governance", "documentation", "security_gates" ], "moe_experts": [ "data_curation_expert", "governance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
f268a2fd4dc135af481794d30b9a00e0a5fc40bf
train_20244
2026-01-01T00:00:00
Latency, cost, and reliability optimization
data_pipeline
advanced
Task: data_pipeline Topic: Latency, cost, and reliability optimization Difficulty: advanced Target language: C# Context: Offline/local deployment with limited compute. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Pipeline: Ingest → Normalize → Filter → Dedupe → Quality → Mix → Audit
[]
{ "target_language": "C#", "developer_needs": [ "evaluation_metrics", "security_gates", "reproducibility", "tooling" ], "moe_experts": [ "evaluation_expert", "data_curation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
17e269b3a33a448236f9b9e0c2a79081f8104ae9
train_20245
2026-01-01T00:00:00
Governance, provenance, and licensing for code data
code
intermediate
Task: code Topic: Governance, provenance, and licensing for code data Difficulty: intermediate Target language: Rust Context: Large monorepo with flaky tests and strict CI. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Design with risks, metrics, acceptance criteria
[]
{ "target_language": "Rust", "developer_needs": [ "evaluation_metrics", "reproducibility", "tests_are_truth", "ci_integration" ], "moe_experts": [ "performance_expert", "data_curation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
8a2b49d7d04c4f2b0a7842e1d4cbefb167414090
train_20246
2026-01-01T00:00:00
Code-specialized model families and sizing tradeoffs
review
expert
Task: review Topic: Code-specialized model families and sizing tradeoffs Difficulty: expert Target language: Bash Context: Regulated environment requiring audit trails. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Review: correctness, security, performance, governance
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "Bash", "developer_needs": [ "reproducibility", "auditability", "ci_integration", "documentation" ], "moe_experts": [ "data_curation_expert", "governance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
35ca6a4eb258d972894e5e0bbbc01760db157d42
train_20247
2026-01-01T00:00:00
Dataset curation pipelines (filter, dedupe, quality)
review
advanced
Task: review Topic: Dataset curation pipelines (filter, dedupe, quality) Difficulty: advanced Target language: Rust Context: High-traffic service with latency SLOs. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Review: correctness, security, performance, governance
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "Rust", "developer_needs": [ "repo_scale_reasoning", "security_gates", "governance", "cost_latency_tradeoffs" ], "moe_experts": [ "evaluation_expert", "governance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
099c97f6176f0a0b2105f9d5575c535a765ef9b1
train_20248
2026-01-01T00:00:00
Governance, provenance, and licensing for code data
review
expert
Task: review Topic: Governance, provenance, and licensing for code data Difficulty: expert Target language: Go Context: Large monorepo with flaky tests and strict CI. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Review: correctness, security, performance, governance
[]
{ "target_language": "Go", "developer_needs": [ "security_gates", "cost_latency_tradeoffs", "reproducibility", "documentation" ], "moe_experts": [ "performance_expert", "governance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
459fa43b6486e1e149c796025265d889b8945fbf
train_20249
2026-01-01T00:00:00
Agentic coding systems (plan→edit→test→reflect)
design
intermediate
Task: design Topic: Agentic coding systems (plan→edit→test→reflect) Difficulty: intermediate Target language: TypeScript Context: Regulated environment requiring audit trails. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Design with risks, metrics, acceptance criteria
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "TypeScript", "developer_needs": [ "documentation", "cost_latency_tradeoffs", "ci_integration", "tests_are_truth" ], "moe_experts": [ "governance_expert", "data_curation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
802431164c1859b38eb8b1c9ae1723545a2c638e
train_20250
2026-01-01T00:00:00
Governance, provenance, and licensing for code data
explain
intermediate
Task: explain Topic: Governance, provenance, and licensing for code data Difficulty: intermediate Target language: C# Context: High-traffic service with latency SLOs. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Design with risks, metrics, acceptance criteria
[]
{ "target_language": "C#", "developer_needs": [ "cost_latency_tradeoffs", "reproducibility", "evaluation_metrics", "tooling" ], "moe_experts": [ "security_expert", "data_curation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
9e9a6ca9d19c7b9f57ffcf8ce794740b0ffa1080
train_20251
2026-01-01T00:00:00
Governance, provenance, and licensing for code data
patch_diff
advanced
Task: patch_diff Topic: Governance, provenance, and licensing for code data Difficulty: advanced Target language: TypeScript Context: Offline/local deployment with limited compute. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Patch (diff-style): ```diff - if x == 0: - return 1/x + if x == 0: + raise ValueError('division by zero') ``` Acceptance: - Tests pass - No new regressions
[]
{ "target_language": "TypeScript", "developer_needs": [ "governance", "cost_latency_tradeoffs", "tooling", "ci_integration" ], "moe_experts": [ "governance_expert", "data_curation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
d8670833baf0802203f1650142d9d98e3fc4045a
train_20252
2026-01-01T00:00:00
Mixture-of-Experts (MoE) for code
patch_diff
advanced
Task: patch_diff Topic: Mixture-of-Experts (MoE) for code Difficulty: advanced Target language: C# Context: Research team validating claims against real repos. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Patch (diff-style): ```diff - if x == 0: - return 1/x + if x == 0: + raise ValueError('division by zero') ``` Acceptance: - Tests pass - No new regressions
[]
{ "target_language": "C#", "developer_needs": [ "auditability", "evaluation_metrics", "documentation", "governance" ], "moe_experts": [ "agentic_systems_expert", "data_curation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
1137d3adce98b56bee53c9205d61849703415b1e
train_20253
2026-01-01T00:00:00
Latency, cost, and reliability optimization
agent_loop
expert
Task: agent_loop Topic: Latency, cost, and reliability optimization Difficulty: expert Target language: C# Context: Research team validating claims against real repos. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Loop: Plan → Edit → Test → Reflect → Human gate
[]
{ "target_language": "C#", "developer_needs": [ "tests_are_truth", "auditability", "repo_scale_reasoning", "security_gates" ], "moe_experts": [ "performance_expert", "coding_expert" ], "governance": { "audit_required": true, "tests_required": true } }
d41742ac6bc3fee5ab93cd19f10ca0d2cf94c3ae
train_20254
2026-01-01T00:00:00
Code-specialized model families and sizing tradeoffs
failure_analysis
intermediate
Task: failure_analysis Topic: Code-specialized model families and sizing tradeoffs Difficulty: intermediate Target language: SQL Context: High-traffic service with latency SLOs. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Failure: - Initial patch broke edge case Reflection: - Missing zero-input guard Correction: - Add explicit validation + test
[]
{ "target_language": "SQL", "developer_needs": [ "security_gates", "auditability", "governance", "cost_latency_tradeoffs" ], "moe_experts": [ "security_expert", "coding_expert" ], "governance": { "audit_required": true, "tests_required": true } }
e2aa0c93d623f387c7126a54a14057fa7e268b11
train_20255
2026-01-01T00:00:00
Code-specialized model families and sizing tradeoffs
explain
expert
Task: explain Topic: Code-specialized model families and sizing tradeoffs Difficulty: expert Target language: Go Context: High-traffic service with latency SLOs. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Design with risks, metrics, acceptance criteria
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "Go", "developer_needs": [ "security_gates", "auditability", "evaluation_metrics", "reproducibility" ], "moe_experts": [ "performance_expert", "security_expert" ], "governance": { "audit_required": true, "tests_required": true } }
2323b12110dee14f3e0370e774e9f2de3013ea60
train_20256
2026-01-01T00:00:00
Dataset curation pipelines (filter, dedupe, quality)
agent_loop
expert
Task: agent_loop Topic: Dataset curation pipelines (filter, dedupe, quality) Difficulty: expert Target language: C# Context: Research team validating claims against real repos. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Loop: Plan → Edit → Test → Reflect → Human gate
[]
{ "target_language": "C#", "developer_needs": [ "evaluation_metrics", "cost_latency_tradeoffs", "ci_integration", "auditability" ], "moe_experts": [ "security_expert", "coding_expert" ], "governance": { "audit_required": true, "tests_required": true } }
10e6bc6d048e41a011fa69c2b8d167c614e62b46
train_20257
2026-01-01T00:00:00
Dataset curation pipelines (filter, dedupe, quality)
compare
advanced
Task: compare Topic: Dataset curation pipelines (filter, dedupe, quality) Difficulty: advanced Target language: Rust Context: Regulated environment requiring audit trails. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Compare: capability, cost, latency, reliability
[]
{ "target_language": "Rust", "developer_needs": [ "reproducibility", "security_gates", "tests_are_truth", "auditability" ], "moe_experts": [ "security_expert", "agentic_systems_expert" ], "governance": { "audit_required": true, "tests_required": true } }
b47efe4e499a2cc3f76e4deaf49cf0118350f4f6
train_20258
2026-01-01T00:00:00
Agentic coding systems (plan→edit→test→reflect)
design
advanced
Task: design Topic: Agentic coding systems (plan→edit→test→reflect) Difficulty: advanced Target language: JavaScript Context: High-traffic service with latency SLOs. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Design with risks, metrics, acceptance criteria
[]
{ "target_language": "JavaScript", "developer_needs": [ "governance", "ci_integration", "tests_are_truth", "repo_scale_reasoning" ], "moe_experts": [ "security_expert", "coding_expert" ], "governance": { "audit_required": true, "tests_required": true } }
ee85eabcee6e4074a97bfb4395155cdbb0c57c65
train_20259
2026-01-01T00:00:00
Dataset curation pipelines (filter, dedupe, quality)
compare
expert
Task: compare Topic: Dataset curation pipelines (filter, dedupe, quality) Difficulty: expert Target language: SQL Context: High-traffic service with latency SLOs. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Compare: capability, cost, latency, reliability
[]
{ "target_language": "SQL", "developer_needs": [ "governance", "tests_are_truth", "auditability", "cost_latency_tradeoffs" ], "moe_experts": [ "agentic_systems_expert", "coding_expert" ], "governance": { "audit_required": true, "tests_required": true } }
8b59d1bc448e936582d93e2a782a68cdbc4e3068
train_20260
2026-01-01T00:00:00
Agentic coding systems (plan→edit→test→reflect)
eval
expert
Task: eval Topic: Agentic coding systems (plan→edit→test→reflect) Difficulty: expert Target language: Java Context: Large monorepo with flaky tests and strict CI. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Eval: pass@k, time-to-green, regressions, diff size
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "Java", "developer_needs": [ "security_gates", "auditability", "tests_are_truth", "cost_latency_tradeoffs" ], "moe_experts": [ "performance_expert", "evaluation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
3902d7de1c87f7b27556b4bdb5c09e70ecc2e908
train_20261
2026-01-01T00:00:00
Mixture-of-Experts (MoE) for code
explain
expert
Task: explain Topic: Mixture-of-Experts (MoE) for code Difficulty: expert Target language: Java Context: Regulated environment requiring audit trails. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Design with risks, metrics, acceptance criteria
[]
{ "target_language": "Java", "developer_needs": [ "tooling", "cost_latency_tradeoffs", "reproducibility", "evaluation_metrics" ], "moe_experts": [ "evaluation_expert", "performance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
3e44c8cdf9dbf638d9de5131683f9a92d4fa88d7
train_20262
2026-01-01T00:00:00
Model merging, distillation, and continued pretraining
code
intermediate
Task: code Topic: Model merging, distillation, and continued pretraining Difficulty: intermediate Target language: TypeScript Context: Offline/local deployment with limited compute. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Design with risks, metrics, acceptance criteria
[]
{ "target_language": "TypeScript", "developer_needs": [ "cost_latency_tradeoffs", "governance", "tests_are_truth", "security_gates" ], "moe_experts": [ "data_curation_expert", "performance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
566f3c300bfa13159ba617e595c7eec4b2d8a1fc
train_20263
2026-01-01T00:00:00
Self-improving agents and feedback loops
data_pipeline
expert
Task: data_pipeline Topic: Self-improving agents and feedback loops Difficulty: expert Target language: Python Context: High-traffic service with latency SLOs. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Pipeline: Ingest → Normalize → Filter → Dedupe → Quality → Mix → Audit
[]
{ "target_language": "Python", "developer_needs": [ "tests_are_truth", "documentation", "governance", "ci_integration" ], "moe_experts": [ "performance_expert", "security_expert" ], "governance": { "audit_required": true, "tests_required": true } }
9b1de3a50b7f3c7ab2d848d17b17bc06cc84d77d
train_20264
2026-01-01T00:00:00
Secure code generation and policy gates
data_pipeline
intermediate
Task: data_pipeline Topic: Secure code generation and policy gates Difficulty: intermediate Target language: SQL Context: Offline/local deployment with limited compute. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Pipeline: Ingest → Normalize → Filter → Dedupe → Quality → Mix → Audit
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "SQL", "developer_needs": [ "repo_scale_reasoning", "reproducibility", "tests_are_truth", "tooling" ], "moe_experts": [ "agentic_systems_expert", "data_curation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
ce0b8f38d460f94ca5246a6f24e41b0b1c5af70d
train_20265
2026-01-01T00:00:00
Latency, cost, and reliability optimization
compare
expert
Task: compare Topic: Latency, cost, and reliability optimization Difficulty: expert Target language: Java Context: Large monorepo with flaky tests and strict CI. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Compare: capability, cost, latency, reliability
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "Java", "developer_needs": [ "governance", "security_gates", "documentation", "cost_latency_tradeoffs" ], "moe_experts": [ "evaluation_expert", "performance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
00b9a6999679bdcd3c0a811b2eb4a04a203069bf
train_20266
2026-01-01T00:00:00
SWE-bench style real-repo evaluation
patch_diff
advanced
Task: patch_diff Topic: SWE-bench style real-repo evaluation Difficulty: advanced Target language: TypeScript Context: High-traffic service with latency SLOs. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Patch (diff-style): ```diff - if x == 0: - return 1/x + if x == 0: + raise ValueError('division by zero') ``` Acceptance: - Tests pass - No new regressions
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "TypeScript", "developer_needs": [ "documentation", "cost_latency_tradeoffs", "governance", "tests_are_truth" ], "moe_experts": [ "coding_expert", "agentic_systems_expert" ], "governance": { "audit_required": true, "tests_required": true } }
0690c7c6aba656b7c6316df3ac973a5ad3fb1ec3
train_20267
2026-01-01T00:00:00
SWE-bench style real-repo evaluation
design
expert
Task: design Topic: SWE-bench style real-repo evaluation Difficulty: expert Target language: Go Context: Research team validating claims against real repos. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Design with risks, metrics, acceptance criteria
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "Go", "developer_needs": [ "cost_latency_tradeoffs", "ci_integration", "repo_scale_reasoning", "tests_are_truth" ], "moe_experts": [ "data_curation_expert", "agentic_systems_expert" ], "governance": { "audit_required": true, "tests_required": true } }
b7aa6e4c72b46fde7802a5913a1340fba147be46
train_20268
2026-01-01T00:00:00
Governance, provenance, and licensing for code data
code
intermediate
Task: code Topic: Governance, provenance, and licensing for code data Difficulty: intermediate Target language: Rust Context: High-traffic service with latency SLOs. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Design with risks, metrics, acceptance criteria
[]
{ "target_language": "Rust", "developer_needs": [ "auditability", "cost_latency_tradeoffs", "governance", "documentation" ], "moe_experts": [ "evaluation_expert", "agentic_systems_expert" ], "governance": { "audit_required": true, "tests_required": true } }
603fe480aa500b36094a63da7b7f0097de0a4f31
train_20269
2026-01-01T00:00:00
Agentic coding systems (plan→edit→test→reflect)
code
intermediate
Task: code Topic: Agentic coding systems (plan→edit→test→reflect) Difficulty: intermediate Target language: SQL Context: High-traffic service with latency SLOs. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Design with risks, metrics, acceptance criteria
[]
{ "target_language": "SQL", "developer_needs": [ "tests_are_truth", "repo_scale_reasoning", "tooling", "reproducibility" ], "moe_experts": [ "coding_expert", "data_curation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
c407232f486ee1d91866f59f3ad7d7f54eb21dd2
train_20270
2026-01-01T00:00:00
Self-improving agents and feedback loops
data_pipeline
advanced
Task: data_pipeline Topic: Self-improving agents and feedback loops Difficulty: advanced Target language: C# Context: Large monorepo with flaky tests and strict CI. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Pipeline: Ingest → Normalize → Filter → Dedupe → Quality → Mix → Audit
[]
{ "target_language": "C#", "developer_needs": [ "tooling", "evaluation_metrics", "tests_are_truth", "repo_scale_reasoning" ], "moe_experts": [ "coding_expert", "agentic_systems_expert" ], "governance": { "audit_required": true, "tests_required": true } }
2ed4f64058de4833e897a20725576b0f58ef5bb3
train_20271
2026-01-01T00:00:00
Multimodal dev workflows (docs, diagrams, traces)
code
intermediate
Task: code Topic: Multimodal dev workflows (docs, diagrams, traces) Difficulty: intermediate Target language: Bash Context: Offline/local deployment with limited compute. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Design with risks, metrics, acceptance criteria
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "Bash", "developer_needs": [ "reproducibility", "evaluation_metrics", "auditability", "ci_integration" ], "moe_experts": [ "governance_expert", "data_curation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
f00da9594f3e5f1360450074e4123c8b970b774e
train_20272
2026-01-01T00:00:00
Code-specialized model families and sizing tradeoffs
agent_loop
expert
Task: agent_loop Topic: Code-specialized model families and sizing tradeoffs Difficulty: expert Target language: Bash Context: Offline/local deployment with limited compute. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Loop: Plan → Edit → Test → Reflect → Human gate
[]
{ "target_language": "Bash", "developer_needs": [ "security_gates", "evaluation_metrics", "tooling", "auditability" ], "moe_experts": [ "data_curation_expert", "performance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
9eb4fd6514873538b5f1ca8ea4005096a20c1018
train_20273
2026-01-01T00:00:00
Agentic coding systems (plan→edit→test→reflect)
compare
expert
Task: compare Topic: Agentic coding systems (plan→edit→test→reflect) Difficulty: expert Target language: Java Context: Research team validating claims against real repos. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Compare: capability, cost, latency, reliability
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "Java", "developer_needs": [ "reproducibility", "ci_integration", "cost_latency_tradeoffs", "repo_scale_reasoning" ], "moe_experts": [ "governance_expert", "evaluation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
1b63a964afb1cd5939bca4b2258af7bebd774881
train_20274
2026-01-01T00:00:00
Self-improving agents and feedback loops
agent_loop
expert
Task: agent_loop Topic: Self-improving agents and feedback loops Difficulty: expert Target language: Python Context: Research team validating claims against real repos. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Loop: Plan → Edit → Test → Reflect → Human gate
[]
{ "target_language": "Python", "developer_needs": [ "cost_latency_tradeoffs", "tooling", "governance", "auditability" ], "moe_experts": [ "agentic_systems_expert", "governance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
1d2b8f0b43b7ce62ca4874c331492866b8637ed2
train_20275
2026-01-01T00:00:00
Self-improving agents and feedback loops
eval
intermediate
Task: eval Topic: Self-improving agents and feedback loops Difficulty: intermediate Target language: C# Context: Research team validating claims against real repos. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Eval: pass@k, time-to-green, regressions, diff size
[]
{ "target_language": "C#", "developer_needs": [ "reproducibility", "auditability", "ci_integration", "repo_scale_reasoning" ], "moe_experts": [ "agentic_systems_expert", "evaluation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
288f29b1bd60e76339565134f7bd1c98f03ed2b2
train_20276
2026-01-01T00:00:00
Model merging, distillation, and continued pretraining
review
advanced
Task: review Topic: Model merging, distillation, and continued pretraining Difficulty: advanced Target language: Bash Context: Research team validating claims against real repos. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Review: correctness, security, performance, governance
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "Bash", "developer_needs": [ "governance", "reproducibility", "ci_integration", "tooling" ], "moe_experts": [ "security_expert", "performance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
e2e608dac3ce1b5289b795662b529301ca3b9f42
train_20277
2026-01-01T00:00:00
Mixture-of-Experts (MoE) for code
data_pipeline
intermediate
Task: data_pipeline Topic: Mixture-of-Experts (MoE) for code Difficulty: intermediate Target language: Bash Context: Offline/local deployment with limited compute. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Pipeline: Ingest → Normalize → Filter → Dedupe → Quality → Mix → Audit
[]
{ "target_language": "Bash", "developer_needs": [ "repo_scale_reasoning", "evaluation_metrics", "governance", "tests_are_truth" ], "moe_experts": [ "data_curation_expert", "performance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
cf745f55acbaa29f2753731e4d03e7a4d13a9b53
train_20278
2026-01-01T00:00:00
Tool calling, sandboxes, and CI integration
explain
expert
Task: explain Topic: Tool calling, sandboxes, and CI integration Difficulty: expert Target language: C# Context: Offline/local deployment with limited compute. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Design with risks, metrics, acceptance criteria
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "C#", "developer_needs": [ "governance", "tooling", "ci_integration", "repo_scale_reasoning" ], "moe_experts": [ "coding_expert", "security_expert" ], "governance": { "audit_required": true, "tests_required": true } }
3467ad3ed3bfaedf01e1f07fd00d47a07b73b580
train_20279
2026-01-01T00:00:00
Self-improving agents and feedback loops
eval
expert
Task: eval Topic: Self-improving agents and feedback loops Difficulty: expert Target language: Go Context: Regulated environment requiring audit trails. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Eval: pass@k, time-to-green, regressions, diff size
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "Go", "developer_needs": [ "reproducibility", "tooling", "governance", "ci_integration" ], "moe_experts": [ "performance_expert", "data_curation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
498a9d816c69025ffcf443fe9e044b44a0d31c0e
train_20280
2026-01-01T00:00:00
Self-improving agents and feedback loops
agent_loop
expert
Task: agent_loop Topic: Self-improving agents and feedback loops Difficulty: expert Target language: Go Context: Research team validating claims against real repos. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Loop: Plan → Edit → Test → Reflect → Human gate
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "Go", "developer_needs": [ "auditability", "repo_scale_reasoning", "documentation", "ci_integration" ], "moe_experts": [ "security_expert", "coding_expert" ], "governance": { "audit_required": true, "tests_required": true } }
eda990850506befedd03543b52d514f557db89af
train_20281
2026-01-01T00:00:00
Self-improving agents and feedback loops
code
intermediate
Task: code Topic: Self-improving agents and feedback loops Difficulty: intermediate Target language: Java Context: Offline/local deployment with limited compute. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Design with risks, metrics, acceptance criteria
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "Java", "developer_needs": [ "security_gates", "auditability", "tooling", "ci_integration" ], "moe_experts": [ "governance_expert", "evaluation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
5935edfd6355b4a0e1f318b19efb2a20a2b7de3c
train_20282
2026-01-01T00:00:00
Tool calling, sandboxes, and CI integration
agent_loop
expert
Task: agent_loop Topic: Tool calling, sandboxes, and CI integration Difficulty: expert Target language: Rust Context: Offline/local deployment with limited compute. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Loop: Plan → Edit → Test → Reflect → Human gate
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "Rust", "developer_needs": [ "repo_scale_reasoning", "security_gates", "auditability", "documentation" ], "moe_experts": [ "governance_expert", "coding_expert" ], "governance": { "audit_required": true, "tests_required": true } }
53014c05562c37fb24910a9532b3fa92e72a263a
train_20283
2026-01-01T00:00:00
Dataset curation pipelines (filter, dedupe, quality)
failure_analysis
advanced
Task: failure_analysis Topic: Dataset curation pipelines (filter, dedupe, quality) Difficulty: advanced Target language: Java Context: Regulated environment requiring audit trails. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Failure: - Initial patch broke edge case Reflection: - Missing zero-input guard Correction: - Add explicit validation + test
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "Java", "developer_needs": [ "security_gates", "cost_latency_tradeoffs", "evaluation_metrics", "tooling" ], "moe_experts": [ "agentic_systems_expert", "coding_expert" ], "governance": { "audit_required": true, "tests_required": true } }
916122beb4c6745ce2bcccfc4e7a518c7a54ce79
train_20284
2026-01-01T00:00:00
Multimodal dev workflows (docs, diagrams, traces)
compare
expert
Task: compare Topic: Multimodal dev workflows (docs, diagrams, traces) Difficulty: expert Target language: JavaScript Context: Large monorepo with flaky tests and strict CI. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Compare: capability, cost, latency, reliability
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "JavaScript", "developer_needs": [ "tooling", "documentation", "reproducibility", "cost_latency_tradeoffs" ], "moe_experts": [ "performance_expert", "coding_expert" ], "governance": { "audit_required": true, "tests_required": true } }
a0c82154a6f2b784b896dc614f11f348f921d13d
train_20285
2026-01-01T00:00:00
Secure code generation and policy gates
patch_diff
advanced
Task: patch_diff Topic: Secure code generation and policy gates Difficulty: advanced Target language: C# Context: Regulated environment requiring audit trails. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Patch (diff-style): ```diff - if x == 0: - return 1/x + if x == 0: + raise ValueError('division by zero') ``` Acceptance: - Tests pass - No new regressions
[]
{ "target_language": "C#", "developer_needs": [ "cost_latency_tradeoffs", "auditability", "tooling", "evaluation_metrics" ], "moe_experts": [ "evaluation_expert", "coding_expert" ], "governance": { "audit_required": true, "tests_required": true } }
37e4e32d5e9ffb16143f34c28e778a3bdaab6957
train_20286
2026-01-01T00:00:00
Agentic coding systems (plan→edit→test→reflect)
failure_analysis
expert
Task: failure_analysis Topic: Agentic coding systems (plan→edit→test→reflect) Difficulty: expert Target language: Bash Context: Regulated environment requiring audit trails. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Failure: - Initial patch broke edge case Reflection: - Missing zero-input guard Correction: - Add explicit validation + test
[]
{ "target_language": "Bash", "developer_needs": [ "repo_scale_reasoning", "documentation", "auditability", "cost_latency_tradeoffs" ], "moe_experts": [ "coding_expert", "evaluation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
fe5d39819828f03342663ab457f22e446597aed2
train_20287
2026-01-01T00:00:00
SWE-bench style real-repo evaluation
code
expert
Task: code Topic: SWE-bench style real-repo evaluation Difficulty: expert Target language: TypeScript Context: Regulated environment requiring audit trails. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Design with risks, metrics, acceptance criteria
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "TypeScript", "developer_needs": [ "documentation", "security_gates", "cost_latency_tradeoffs", "tests_are_truth" ], "moe_experts": [ "governance_expert", "coding_expert" ], "governance": { "audit_required": true, "tests_required": true } }
0ef8b16abf2c38c0544b82ab54c2d06c6ba667c6
train_20288
2026-01-01T00:00:00
Self-improving agents and feedback loops
compare
intermediate
Task: compare Topic: Self-improving agents and feedback loops Difficulty: intermediate Target language: JavaScript Context: Research team validating claims against real repos. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Compare: capability, cost, latency, reliability
[]
{ "target_language": "JavaScript", "developer_needs": [ "repo_scale_reasoning", "security_gates", "cost_latency_tradeoffs", "reproducibility" ], "moe_experts": [ "governance_expert", "data_curation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
498453613423d9a197c377119b88731ef1f32b91
train_20289
2026-01-01T00:00:00
SWE-bench style real-repo evaluation
compare
expert
Task: compare Topic: SWE-bench style real-repo evaluation Difficulty: expert Target language: Java Context: Offline/local deployment with limited compute. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Compare: capability, cost, latency, reliability
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "Java", "developer_needs": [ "reproducibility", "tests_are_truth", "security_gates", "governance" ], "moe_experts": [ "data_curation_expert", "governance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
3d74931321744b25fecff8b48cb56f4857b3620a
train_20290
2026-01-01T00:00:00
Self-improving agents and feedback loops
review
intermediate
Task: review Topic: Self-improving agents and feedback loops Difficulty: intermediate Target language: C# Context: High-traffic service with latency SLOs. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Review: correctness, security, performance, governance
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "C#", "developer_needs": [ "documentation", "cost_latency_tradeoffs", "tests_are_truth", "tooling" ], "moe_experts": [ "performance_expert", "governance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
5f9a1436f57856a61f06cc52bf88d4f2599e0afa
train_20291
2026-01-01T00:00:00
Reasoning-first coding models and tunable deliberation
eval
advanced
Task: eval Topic: Reasoning-first coding models and tunable deliberation Difficulty: advanced Target language: Go Context: Research team validating claims against real repos. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Eval: pass@k, time-to-green, regressions, diff size
[]
{ "target_language": "Go", "developer_needs": [ "documentation", "auditability", "cost_latency_tradeoffs", "ci_integration" ], "moe_experts": [ "governance_expert", "agentic_systems_expert" ], "governance": { "audit_required": true, "tests_required": true } }
2fe5a0d05aa7ee58cbadc84e45545526977c2187
train_20292
2026-01-01T00:00:00
Secure code generation and policy gates
review
advanced
Task: review Topic: Secure code generation and policy gates Difficulty: advanced Target language: C# Context: Regulated environment requiring audit trails. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Review: correctness, security, performance, governance
[]
{ "target_language": "C#", "developer_needs": [ "reproducibility", "security_gates", "governance", "tests_are_truth" ], "moe_experts": [ "performance_expert", "governance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
99b1e450d6a38cf00afa7f4d9d9cb1cd472b2544
train_20293
2026-01-01T00:00:00
Tool calling, sandboxes, and CI integration
code
advanced
Task: code Topic: Tool calling, sandboxes, and CI integration Difficulty: advanced Target language: TypeScript Context: Large monorepo with flaky tests and strict CI. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Design with risks, metrics, acceptance criteria
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "TypeScript", "developer_needs": [ "documentation", "repo_scale_reasoning", "tests_are_truth", "cost_latency_tradeoffs" ], "moe_experts": [ "performance_expert", "agentic_systems_expert" ], "governance": { "audit_required": true, "tests_required": true } }
f684ee0a2c2f9056dba79053986f7edbc25b9705
train_20294
2026-01-01T00:00:00
Latency, cost, and reliability optimization
eval
advanced
Task: eval Topic: Latency, cost, and reliability optimization Difficulty: advanced Target language: Java Context: Large monorepo with flaky tests and strict CI. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Eval: pass@k, time-to-green, regressions, diff size
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "Java", "developer_needs": [ "repo_scale_reasoning", "reproducibility", "auditability", "governance" ], "moe_experts": [ "performance_expert", "governance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
97e8a630067669175c2a0d047ceb6b9494d8fa27
train_20295
2026-01-01T00:00:00
Code-specialized model families and sizing tradeoffs
agent_loop
intermediate
Task: agent_loop Topic: Code-specialized model families and sizing tradeoffs Difficulty: intermediate Target language: Go Context: Research team validating claims against real repos. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Loop: Plan → Edit → Test → Reflect → Human gate
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "Go", "developer_needs": [ "ci_integration", "auditability", "documentation", "evaluation_metrics" ], "moe_experts": [ "performance_expert", "data_curation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
593ad7b4356d14c5e073294b34b0210dba583488
train_20296
2026-01-01T00:00:00
Code-specialized model families and sizing tradeoffs
design
intermediate
Task: design Topic: Code-specialized model families and sizing tradeoffs Difficulty: intermediate Target language: Go Context: High-traffic service with latency SLOs. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Design with risks, metrics, acceptance criteria
[]
{ "target_language": "Go", "developer_needs": [ "auditability", "tooling", "security_gates", "ci_integration" ], "moe_experts": [ "governance_expert", "evaluation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
c41419c5eba8fa123e46e4e6524c739b7fc6a555
train_20297
2026-01-01T00:00:00
Extended context and repo-scale understanding
code
expert
Task: code Topic: Extended context and repo-scale understanding Difficulty: expert Target language: JavaScript Context: Offline/local deployment with limited compute. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Design with risks, metrics, acceptance criteria
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "JavaScript", "developer_needs": [ "security_gates", "auditability", "repo_scale_reasoning", "documentation" ], "moe_experts": [ "data_curation_expert", "coding_expert" ], "governance": { "audit_required": true, "tests_required": true } }
79504825d188add483d82f394c7e313fb86367f9
train_20298
2026-01-01T00:00:00
Dataset curation pipelines (filter, dedupe, quality)
design
intermediate
Task: design Topic: Dataset curation pipelines (filter, dedupe, quality) Difficulty: intermediate Target language: Go Context: High-traffic service with latency SLOs. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Design with risks, metrics, acceptance criteria
[]
{ "target_language": "Go", "developer_needs": [ "tooling", "governance", "evaluation_metrics", "tests_are_truth" ], "moe_experts": [ "performance_expert", "security_expert" ], "governance": { "audit_required": true, "tests_required": true } }
fc7771e8ef8e9d8f6309e06df1b82e422593818f
train_20299
2026-01-01T00:00:00
SWE-bench style real-repo evaluation
patch_diff
intermediate
Task: patch_diff Topic: SWE-bench style real-repo evaluation Difficulty: intermediate Target language: Rust Context: High-traffic service with latency SLOs. Produce expert-level, production-ready artifacts.
Facts: - Modern AI coding prioritizes correctness, evaluation, and governance. - Agentic loops with test gates outperform single-pass generation. Patch (diff-style): ```diff - if x == 0: - return 1/x + if x == 0: + raise ValueError('division by zero') ``` Acceptance: - Tests pass - No new regressions
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "Rust", "developer_needs": [ "tooling", "tests_are_truth", "security_gates", "auditability" ], "moe_experts": [ "agentic_systems_expert", "performance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
28805155a35d2859e80f2f976ac93a62a9cab644