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_22800
2026-01-01T00:00:00
Dataset curation pipelines (filter, dedupe, quality)
patch_diff
advanced
Task: patch_diff Topic: Dataset curation pipelines (filter, dedupe, quality) 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. 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": [ "auditability", "ci_integration", "repo_scale_reasoning", "tests_are_truth" ], "moe_experts": [ "security_expert", "evaluation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
9af7f6b1dafcd90ee034d9272a915f33ffe2e885
train_22801
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: 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. Design with risks, metrics, acceptance criteria
[]
{ "target_language": "SQL", "developer_needs": [ "governance", "evaluation_metrics", "reproducibility", "documentation" ], "moe_experts": [ "coding_expert", "security_expert" ], "governance": { "audit_required": true, "tests_required": true } }
cc72c192bae800a8215212cc82b57ea7c09ea2a2
train_22802
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: 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. Review: correctness, security, performance, governance
[]
{ "target_language": "C#", "developer_needs": [ "documentation", "ci_integration", "tests_are_truth", "cost_latency_tradeoffs" ], "moe_experts": [ "data_curation_expert", "performance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
0fbf83c4e6e0bc2b06cca3e8cae52ad4d3a4ac2a
train_22803
2026-01-01T00:00:00
Reasoning-first coding models and tunable deliberation
explain
expert
Task: explain Topic: Reasoning-first coding models and tunable deliberation 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. 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": [ "documentation", "cost_latency_tradeoffs", "evaluation_metrics", "security_gates" ], "moe_experts": [ "performance_expert", "coding_expert" ], "governance": { "audit_required": true, "tests_required": true } }
555b67e35006f97b24b5aabc982a1e65262fc047
train_22804
2026-01-01T00:00:00
SWE-bench style real-repo evaluation
explain
advanced
Task: explain Topic: SWE-bench style real-repo evaluation Difficulty: advanced 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": [ "governance", "security_gates", "auditability", "tooling" ], "moe_experts": [ "coding_expert", "evaluation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
f26ab9416b7d5cdc28cab99771262e8f293469ec
train_22805
2026-01-01T00:00:00
Latency, cost, and reliability optimization
review
intermediate
Task: review Topic: Latency, cost, and reliability optimization Difficulty: intermediate 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": [ "auditability", "evaluation_metrics", "documentation", "governance" ], "moe_experts": [ "coding_expert", "agentic_systems_expert" ], "governance": { "audit_required": true, "tests_required": true } }
8a72187bef4a32e9da7b609593101ce7719f3275
train_22806
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: 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
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "Python", "developer_needs": [ "evaluation_metrics", "ci_integration", "repo_scale_reasoning", "cost_latency_tradeoffs" ], "moe_experts": [ "evaluation_expert", "agentic_systems_expert" ], "governance": { "audit_required": true, "tests_required": true } }
8577b127544c50f9d885ea54f68dc407930425e7
train_22807
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: 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. Design with risks, metrics, acceptance criteria
[]
{ "target_language": "Bash", "developer_needs": [ "tests_are_truth", "cost_latency_tradeoffs", "reproducibility", "repo_scale_reasoning" ], "moe_experts": [ "performance_expert", "agentic_systems_expert" ], "governance": { "audit_required": true, "tests_required": true } }
3995765778210767e00f68632e4309d4b98141f0
train_22808
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: 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": [ "security_gates", "tooling", "repo_scale_reasoning", "auditability" ], "moe_experts": [ "data_curation_expert", "coding_expert" ], "governance": { "audit_required": true, "tests_required": true } }
a82c04c457636f32d1be3e4eb7c3d6db287c9476
train_22809
2026-01-01T00:00:00
Reasoning-first coding models and tunable deliberation
data_pipeline
expert
Task: data_pipeline Topic: Reasoning-first coding models and tunable deliberation 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. 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": [ "tooling", "evaluation_metrics", "reproducibility", "cost_latency_tradeoffs" ], "moe_experts": [ "coding_expert", "governance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
8f6b9093a92d9970e967ce6eee8cbebe7e607874
train_22810
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: 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": [ "governance", "cost_latency_tradeoffs", "documentation", "tooling" ], "moe_experts": [ "governance_expert", "data_curation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
70b11b84be0213249f39ff2d20b2e07372c4a91f
train_22811
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: 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. 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": [ "governance", "cost_latency_tradeoffs", "ci_integration", "security_gates" ], "moe_experts": [ "agentic_systems_expert", "coding_expert" ], "governance": { "audit_required": true, "tests_required": true } }
a9e65f5881bf33b51c85e5c37f489b7d544a4327
train_22812
2026-01-01T00:00:00
Dataset curation pipelines (filter, dedupe, quality)
failure_analysis
intermediate
Task: failure_analysis Topic: Dataset curation pipelines (filter, dedupe, quality) Difficulty: intermediate 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. 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": "Bash", "developer_needs": [ "evaluation_metrics", "ci_integration", "reproducibility", "documentation" ], "moe_experts": [ "coding_expert", "evaluation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
d39fd726e37eaa3fb985f89d4ff619c5b3d9f1e3
train_22813
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: 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
[]
{ "target_language": "JavaScript", "developer_needs": [ "security_gates", "repo_scale_reasoning", "reproducibility", "documentation" ], "moe_experts": [ "performance_expert", "data_curation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
117d5aa3bb53de119878b3ebc8efdf8727a4b2d4
train_22814
2026-01-01T00:00:00
Tool calling, sandboxes, and CI integration
data_pipeline
advanced
Task: data_pipeline Topic: Tool calling, sandboxes, and CI integration Difficulty: advanced 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": [ "tests_are_truth", "reproducibility", "tooling", "documentation" ], "moe_experts": [ "security_expert", "coding_expert" ], "governance": { "audit_required": true, "tests_required": true } }
0bae9704113f2be2322c1d11f705c59e539ccce3
train_22815
2026-01-01T00:00:00
Agentic coding systems (plan→edit→test→reflect)
compare
intermediate
Task: compare Topic: Agentic coding systems (plan→edit→test→reflect) Difficulty: intermediate 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. 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": [ "security_gates", "reproducibility", "tooling", "tests_are_truth" ], "moe_experts": [ "governance_expert", "security_expert" ], "governance": { "audit_required": true, "tests_required": true } }
981cf04e5e76c538f039a3a2ced8503d3db90d09
train_22816
2026-01-01T00:00:00
Tool calling, sandboxes, and CI integration
review
intermediate
Task: review Topic: Tool calling, sandboxes, and CI integration 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. 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": [ "evaluation_metrics", "tests_are_truth", "security_gates", "tooling" ], "moe_experts": [ "governance_expert", "evaluation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
d160c81e40bb990a876632aa751860372af063b7
train_22817
2026-01-01T00:00:00
Mixture-of-Experts (MoE) for code
design
expert
Task: design Topic: Mixture-of-Experts (MoE) for code Difficulty: expert Target language: SQL 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": "SQL", "developer_needs": [ "documentation", "governance", "repo_scale_reasoning", "security_gates" ], "moe_experts": [ "agentic_systems_expert", "evaluation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
c871fe9827c39203af43beb3939732cd7b900c64
train_22818
2026-01-01T00:00:00
Tool calling, sandboxes, and CI integration
design
expert
Task: design Topic: Tool calling, sandboxes, and CI integration 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
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "Go", "developer_needs": [ "repo_scale_reasoning", "security_gates", "evaluation_metrics", "reproducibility" ], "moe_experts": [ "evaluation_expert", "performance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
a3a164268219e8a7381bba7ba7037b9b7cb35243
train_22819
2026-01-01T00:00:00
Extended context and repo-scale understanding
patch_diff
intermediate
Task: patch_diff Topic: Extended context and repo-scale understanding Difficulty: intermediate Target language: Bash 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
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "Bash", "developer_needs": [ "documentation", "cost_latency_tradeoffs", "tooling", "ci_integration" ], "moe_experts": [ "governance_expert", "security_expert" ], "governance": { "audit_required": true, "tests_required": true } }
6c275e4cc3411d402cdc9ce7a689ea02ede5208c
train_22820
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: 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. 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": "JavaScript", "developer_needs": [ "governance", "security_gates", "documentation", "tests_are_truth" ], "moe_experts": [ "agentic_systems_expert", "evaluation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
bb1afe56c68709727227007a8d70f384b4705c22
train_22821
2026-01-01T00:00:00
Agentic coding systems (plan→edit→test→reflect)
data_pipeline
expert
Task: data_pipeline Topic: Agentic coding systems (plan→edit→test→reflect) Difficulty: expert 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. 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": "C#", "developer_needs": [ "tooling", "repo_scale_reasoning", "cost_latency_tradeoffs", "auditability" ], "moe_experts": [ "evaluation_expert", "data_curation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
a00f06c039afc557a7be3a8d8be8490223bfd67b
train_22822
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: 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. 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": "Rust", "developer_needs": [ "ci_integration", "security_gates", "cost_latency_tradeoffs", "governance" ], "moe_experts": [ "data_curation_expert", "agentic_systems_expert" ], "governance": { "audit_required": true, "tests_required": true } }
a53150fc9fea421c3bef73e76f9adf88d5944ca1
train_22823
2026-01-01T00:00:00
Dataset curation pipelines (filter, dedupe, quality)
explain
expert
Task: explain Topic: Dataset curation pipelines (filter, dedupe, quality) Difficulty: expert 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. 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": [ "auditability", "cost_latency_tradeoffs", "repo_scale_reasoning", "evaluation_metrics" ], "moe_experts": [ "performance_expert", "governance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
e622f47194e0036d9be983607368c2707edc872d
train_22824
2026-01-01T00:00:00
Reasoning-first coding models and tunable deliberation
review
expert
Task: review Topic: Reasoning-first coding models and tunable deliberation Difficulty: expert 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. Review: correctness, security, performance, governance
[]
{ "target_language": "Python", "developer_needs": [ "tooling", "evaluation_metrics", "security_gates", "governance" ], "moe_experts": [ "performance_expert", "coding_expert" ], "governance": { "audit_required": true, "tests_required": true } }
8e17c46c2129e33cedbe3da27ee634144f7a3f58
train_22825
2026-01-01T00:00:00
Latency, cost, and reliability optimization
design
intermediate
Task: design Topic: Latency, cost, and reliability optimization Difficulty: intermediate 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": [ "evaluation_metrics", "repo_scale_reasoning", "cost_latency_tradeoffs", "auditability" ], "moe_experts": [ "performance_expert", "evaluation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
89acd939dc5cf1c5d5c04c85484686cdf5b3fd57
train_22826
2026-01-01T00:00:00
Reasoning-first coding models and tunable deliberation
agent_loop
expert
Task: agent_loop Topic: Reasoning-first coding models and tunable deliberation Difficulty: expert 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. 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": "SQL", "developer_needs": [ "cost_latency_tradeoffs", "auditability", "reproducibility", "governance" ], "moe_experts": [ "governance_expert", "evaluation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
5d3c8b50b80176bb3c1b17922401e3c08f48f40b
train_22827
2026-01-01T00:00:00
Dataset curation pipelines (filter, dedupe, quality)
code
intermediate
Task: code Topic: Dataset curation pipelines (filter, dedupe, quality) Difficulty: intermediate 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. Scaffold: ```python def agent_loop(plan, edit, test, issue, max_iters=4): history = [] p = plan(issue) for _ in range(max_iters): patch = edit(issue, p) ok, report = test(patch) history.append({"plan": p, "ok": ok}) if ok: return patch, history p = p + " | refine" return patch, history ```
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "Python", "developer_needs": [ "repo_scale_reasoning", "reproducibility", "evaluation_metrics", "tooling" ], "moe_experts": [ "agentic_systems_expert", "performance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
e08935c44316cf992f7d08743aedf5d09f7609b2
train_22828
2026-01-01T00:00:00
Self-improving agents and feedback loops
compare
expert
Task: compare 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. Compare: capability, cost, latency, reliability
[]
{ "target_language": "Python", "developer_needs": [ "repo_scale_reasoning", "documentation", "reproducibility", "auditability" ], "moe_experts": [ "performance_expert", "evaluation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
cd394ce393bf5f59f0f2906b8d66ca2811b80937
train_22829
2026-01-01T00:00:00
Latency, cost, and reliability optimization
review
intermediate
Task: review Topic: Latency, cost, and reliability optimization 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. Review: correctness, security, performance, governance
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "TypeScript", "developer_needs": [ "tests_are_truth", "security_gates", "cost_latency_tradeoffs", "evaluation_metrics" ], "moe_experts": [ "evaluation_expert", "agentic_systems_expert" ], "governance": { "audit_required": true, "tests_required": true } }
d4912312a2dce020b4c1d6871f64edb53719f63a
train_22830
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: 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. Eval: pass@k, time-to-green, regressions, diff size
[]
{ "target_language": "Bash", "developer_needs": [ "tests_are_truth", "evaluation_metrics", "reproducibility", "documentation" ], "moe_experts": [ "data_curation_expert", "agentic_systems_expert" ], "governance": { "audit_required": true, "tests_required": true } }
f3ceb2932d1c9323f2bca6d496205e1449a62464
train_22831
2026-01-01T00:00:00
SWE-bench style real-repo evaluation
agent_loop
intermediate
Task: agent_loop Topic: SWE-bench style real-repo evaluation 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. Loop: Plan → Edit → Test → Reflect → Human gate
[]
{ "target_language": "Bash", "developer_needs": [ "security_gates", "tooling", "reproducibility", "cost_latency_tradeoffs" ], "moe_experts": [ "data_curation_expert", "performance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
f7e75ff58c5324d75b3962a40bcc1060218aca41
train_22832
2026-01-01T00:00:00
Reasoning-first coding models and tunable deliberation
patch_diff
intermediate
Task: patch_diff Topic: Reasoning-first coding models and tunable deliberation Difficulty: intermediate 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. 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": [ "repo_scale_reasoning", "auditability", "governance", "tooling" ], "moe_experts": [ "evaluation_expert", "coding_expert" ], "governance": { "audit_required": true, "tests_required": true } }
857a18985a94bb9ff8af8adc34a94465a62c649c
train_22833
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: 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. Design with risks, metrics, acceptance criteria
[]
{ "target_language": "TypeScript", "developer_needs": [ "repo_scale_reasoning", "governance", "reproducibility", "tests_are_truth" ], "moe_experts": [ "security_expert", "evaluation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
836f44fa1532d88a0aa593b7af7524f880776d68
train_22834
2026-01-01T00:00:00
Model merging, distillation, and continued pretraining
data_pipeline
advanced
Task: data_pipeline Topic: Model merging, distillation, and continued pretraining Difficulty: advanced 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. Pipeline: Ingest → Normalize → Filter → Dedupe → Quality → Mix → Audit
[]
{ "target_language": "Python", "developer_needs": [ "evaluation_metrics", "ci_integration", "security_gates", "repo_scale_reasoning" ], "moe_experts": [ "evaluation_expert", "security_expert" ], "governance": { "audit_required": true, "tests_required": true } }
c59e0b93d367bd52f8bcd87dd56d5b29c08a0bb0
train_22835
2026-01-01T00:00:00
Latency, cost, and reliability optimization
patch_diff
intermediate
Task: patch_diff Topic: Latency, cost, and reliability optimization Difficulty: intermediate 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. 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": [ "tests_are_truth", "documentation", "governance", "ci_integration" ], "moe_experts": [ "evaluation_expert", "governance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
da253274444462a3a56bbf0acabf9ce6a576f445
train_22836
2026-01-01T00:00:00
Self-improving agents and feedback loops
compare
advanced
Task: compare Topic: Self-improving agents and feedback loops 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. Compare: capability, cost, latency, reliability
[]
{ "target_language": "Rust", "developer_needs": [ "auditability", "security_gates", "evaluation_metrics", "reproducibility" ], "moe_experts": [ "performance_expert", "agentic_systems_expert" ], "governance": { "audit_required": true, "tests_required": true } }
53bf97f86b1ef481548b2510a1414538f889b3b4
train_22837
2026-01-01T00:00:00
Code-specialized model families and sizing tradeoffs
code
expert
Task: code Topic: Code-specialized model families and sizing tradeoffs 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. 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": [ "tests_are_truth", "governance", "tooling", "cost_latency_tradeoffs" ], "moe_experts": [ "coding_expert", "evaluation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
5d6ddcb3bf19a13f8cd102fee0d2c7e6e60d3847
train_22838
2026-01-01T00:00:00
Latency, cost, and reliability optimization
code
expert
Task: code Topic: Latency, cost, and reliability optimization Difficulty: expert Target language: SQL 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": "SQL", "developer_needs": [ "ci_integration", "repo_scale_reasoning", "auditability", "documentation" ], "moe_experts": [ "data_curation_expert", "security_expert" ], "governance": { "audit_required": true, "tests_required": true } }
e3e01ed2c3de34d806b4047a26ffd52448fdacaa
train_22839
2026-01-01T00:00:00
Tool calling, sandboxes, and CI integration
failure_analysis
advanced
Task: failure_analysis Topic: Tool calling, sandboxes, and CI integration Difficulty: advanced Target language: Python 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": "Python", "developer_needs": [ "evaluation_metrics", "tests_are_truth", "reproducibility", "governance" ], "moe_experts": [ "data_curation_expert", "agentic_systems_expert" ], "governance": { "audit_required": true, "tests_required": true } }
92076bfa4879a3599b4aff6ffde635d96d4034ae
train_22840
2026-01-01T00:00:00
SWE-bench style real-repo evaluation
explain
advanced
Task: explain Topic: SWE-bench style real-repo evaluation 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. Design with risks, metrics, acceptance criteria
[]
{ "target_language": "TypeScript", "developer_needs": [ "tests_are_truth", "tooling", "documentation", "reproducibility" ], "moe_experts": [ "performance_expert", "agentic_systems_expert" ], "governance": { "audit_required": true, "tests_required": true } }
59cc8568fc4b2eb715fa7f4e81462f52f3247c9a
train_22841
2026-01-01T00:00:00
Mixture-of-Experts (MoE) for code
patch_diff
intermediate
Task: patch_diff Topic: Mixture-of-Experts (MoE) for code 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. 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": "SQL", "developer_needs": [ "repo_scale_reasoning", "tooling", "governance", "documentation" ], "moe_experts": [ "agentic_systems_expert", "data_curation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
2ae8305c3b7919e956f37dbd3ed81dc7f75dccc6
train_22842
2026-01-01T00:00:00
Agentic coding systems (plan→edit→test→reflect)
data_pipeline
expert
Task: data_pipeline Topic: Agentic coding systems (plan→edit→test→reflect) 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. Pipeline: Ingest → Normalize → Filter → Dedupe → Quality → Mix → Audit
[]
{ "target_language": "TypeScript", "developer_needs": [ "auditability", "documentation", "governance", "evaluation_metrics" ], "moe_experts": [ "governance_expert", "data_curation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
c96b131c334632608d26b76b5cbac113ba2c1a69
train_22843
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: Bash 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": "Bash", "developer_needs": [ "repo_scale_reasoning", "governance", "evaluation_metrics", "tooling" ], "moe_experts": [ "performance_expert", "data_curation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
8c674d767f439b31a8f6738208ec6561cbd63161
train_22844
2026-01-01T00:00:00
Governance, provenance, and licensing for code data
patch_diff
expert
Task: patch_diff Topic: Governance, provenance, and licensing for code data 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. 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": [ "tests_are_truth", "repo_scale_reasoning", "documentation", "evaluation_metrics" ], "moe_experts": [ "governance_expert", "security_expert" ], "governance": { "audit_required": true, "tests_required": true } }
e6830f6d480626b865fe063c298d38aa9581faaa
train_22845
2026-01-01T00:00:00
Agentic coding systems (plan→edit→test→reflect)
patch_diff
expert
Task: patch_diff Topic: Agentic coding systems (plan→edit→test→reflect) Difficulty: expert 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. 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": [ "documentation", "ci_integration", "repo_scale_reasoning", "tests_are_truth" ], "moe_experts": [ "agentic_systems_expert", "governance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
b0d17acae5b00b8d6eea8c5ce14e48b29d98dfe3
train_22846
2026-01-01T00:00:00
Secure code generation and policy gates
failure_analysis
expert
Task: failure_analysis Topic: Secure code generation and policy gates Difficulty: expert 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. Failure: - Initial patch broke edge case Reflection: - Missing zero-input guard Correction: - Add explicit validation + test
[]
{ "target_language": "TypeScript", "developer_needs": [ "tests_are_truth", "repo_scale_reasoning", "tooling", "cost_latency_tradeoffs" ], "moe_experts": [ "security_expert", "data_curation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
27104f9f9ba002dc3348abdaeadd636b02e385c8
train_22847
2026-01-01T00:00:00
Dataset curation pipelines (filter, dedupe, quality)
agent_loop
intermediate
Task: agent_loop Topic: Dataset curation pipelines (filter, dedupe, quality) Difficulty: intermediate 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. Loop: Plan → Edit → Test → Reflect → Human gate
[]
{ "target_language": "Go", "developer_needs": [ "repo_scale_reasoning", "tests_are_truth", "tooling", "auditability" ], "moe_experts": [ "evaluation_expert", "security_expert" ], "governance": { "audit_required": true, "tests_required": true } }
66ba4da29b7fe3eeacb6aa113d36b5c48059ffe6
train_22848
2026-01-01T00:00:00
Extended context and repo-scale understanding
agent_loop
intermediate
Task: agent_loop Topic: Extended context and repo-scale understanding 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. Loop: Plan → Edit → Test → Reflect → Human gate
[]
{ "target_language": "JavaScript", "developer_needs": [ "tooling", "cost_latency_tradeoffs", "governance", "evaluation_metrics" ], "moe_experts": [ "security_expert", "governance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
7b510919f7f3cd3b8759d5d49ddd39e15c94dcb3
train_22849
2026-01-01T00:00:00
Multimodal dev workflows (docs, diagrams, traces)
eval
advanced
Task: eval Topic: Multimodal dev workflows (docs, diagrams, traces) 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. Eval: pass@k, time-to-green, regressions, diff size
[]
{ "target_language": "Java", "developer_needs": [ "tests_are_truth", "repo_scale_reasoning", "evaluation_metrics", "reproducibility" ], "moe_experts": [ "governance_expert", "security_expert" ], "governance": { "audit_required": true, "tests_required": true } }
a8a23a4050ee782d8a3539acfcf807d94fb7820e
train_22850
2026-01-01T00:00:00
Self-improving agents and feedback loops
agent_loop
intermediate
Task: agent_loop Topic: Self-improving agents and feedback loops 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. 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": [ "auditability", "reproducibility", "governance", "ci_integration" ], "moe_experts": [ "data_curation_expert", "security_expert" ], "governance": { "audit_required": true, "tests_required": true } }
cc73c4b27ef6b31b2ec082bfc97635c3a8b349d4
train_22851
2026-01-01T00:00:00
Tool calling, sandboxes, and CI integration
data_pipeline
advanced
Task: data_pipeline Topic: Tool calling, sandboxes, and CI integration 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
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "C#", "developer_needs": [ "auditability", "security_gates", "tooling", "tests_are_truth" ], "moe_experts": [ "security_expert", "coding_expert" ], "governance": { "audit_required": true, "tests_required": true } }
a969d64959ca233183dcd9baddaaa832b62f0739
train_22852
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: 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. Review: correctness, security, performance, governance
[]
{ "target_language": "Go", "developer_needs": [ "security_gates", "governance", "repo_scale_reasoning", "documentation" ], "moe_experts": [ "evaluation_expert", "performance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
0f2560a2f23b0e7cd7c6c423ad44048366d070cb
train_22853
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: 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": [ "security_gates", "ci_integration", "governance", "tests_are_truth" ], "moe_experts": [ "performance_expert", "coding_expert" ], "governance": { "audit_required": true, "tests_required": true } }
cbcb5cbb316d1e02cd9adbae30a1eed5e358ace6
train_22854
2026-01-01T00:00:00
Model merging, distillation, and continued pretraining
data_pipeline
intermediate
Task: data_pipeline Topic: Model merging, distillation, and continued pretraining 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. Pipeline: Ingest → Normalize → Filter → Dedupe → Quality → Mix → Audit
[]
{ "target_language": "Go", "developer_needs": [ "cost_latency_tradeoffs", "ci_integration", "reproducibility", "tests_are_truth" ], "moe_experts": [ "data_curation_expert", "governance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
080982ea5b1ebc0ff7c54c33d881461131c7d6b4
train_22855
2026-01-01T00:00:00
Tool calling, sandboxes, and CI integration
data_pipeline
expert
Task: data_pipeline Topic: Tool calling, sandboxes, and CI integration 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. Pipeline: Ingest → Normalize → Filter → Dedupe → Quality → Mix → Audit
[]
{ "target_language": "TypeScript", "developer_needs": [ "auditability", "security_gates", "governance", "reproducibility" ], "moe_experts": [ "governance_expert", "data_curation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
8e8724f40ec4108f2715c346b172cbf66b996564
train_22856
2026-01-01T00:00:00
Multimodal dev workflows (docs, diagrams, traces)
design
expert
Task: design Topic: Multimodal dev workflows (docs, diagrams, traces) 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
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "SQL", "developer_needs": [ "security_gates", "tooling", "ci_integration", "evaluation_metrics" ], "moe_experts": [ "agentic_systems_expert", "performance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
042ebd0002c0501397376d40132beda486cbe319
train_22857
2026-01-01T00:00:00
Agentic coding systems (plan→edit→test→reflect)
compare
intermediate
Task: compare 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. Compare: capability, cost, latency, reliability
[]
{ "target_language": "Java", "developer_needs": [ "ci_integration", "security_gates", "governance", "evaluation_metrics" ], "moe_experts": [ "coding_expert", "security_expert" ], "governance": { "audit_required": true, "tests_required": true } }
6bf819ba8871a82b9c0ea5738d0323a1cdfebe6c
train_22858
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: 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. Design with risks, metrics, acceptance criteria
[]
{ "target_language": "Go", "developer_needs": [ "documentation", "repo_scale_reasoning", "cost_latency_tradeoffs", "governance" ], "moe_experts": [ "data_curation_expert", "evaluation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
314b0b11efc11918408b5b32a96dceeda4aeb32a
train_22859
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: SQL 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": "SQL", "developer_needs": [ "governance", "evaluation_metrics", "ci_integration", "documentation" ], "moe_experts": [ "agentic_systems_expert", "coding_expert" ], "governance": { "audit_required": true, "tests_required": true } }
504d7a943f184a5635cfcc78f1e2babc010b6541
train_22860
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: 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": [ "auditability", "ci_integration", "cost_latency_tradeoffs", "documentation" ], "moe_experts": [ "coding_expert", "governance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
804be2199a690299ff6ae970600f75f6d7498d64
train_22861
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: 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. 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": "Rust", "developer_needs": [ "documentation", "repo_scale_reasoning", "tooling", "security_gates" ], "moe_experts": [ "governance_expert", "evaluation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
51fe12ec33ba690b357503d4064392ce97e11145
train_22862
2026-01-01T00:00:00
Agentic coding systems (plan→edit→test→reflect)
data_pipeline
advanced
Task: data_pipeline Topic: Agentic coding systems (plan→edit→test→reflect) Difficulty: advanced 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. Pipeline: Ingest → Normalize → Filter → Dedupe → Quality → Mix → Audit
[]
{ "target_language": "Go", "developer_needs": [ "ci_integration", "security_gates", "tooling", "repo_scale_reasoning" ], "moe_experts": [ "coding_expert", "security_expert" ], "governance": { "audit_required": true, "tests_required": true } }
f5f48f8b82febcf7e3e1f5ae3d43c863a5414fca
train_22863
2026-01-01T00:00:00
Latency, cost, and reliability optimization
review
intermediate
Task: review Topic: Latency, cost, and reliability optimization Difficulty: intermediate Target language: Python 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": "Python", "developer_needs": [ "repo_scale_reasoning", "auditability", "security_gates", "documentation" ], "moe_experts": [ "agentic_systems_expert", "evaluation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
dd2bf051c9b9d99c9413d17ef1e393ceb4be08b4
train_22864
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: Bash 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": "Bash", "developer_needs": [ "tooling", "cost_latency_tradeoffs", "evaluation_metrics", "documentation" ], "moe_experts": [ "coding_expert", "data_curation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
4b1fb289bf80467a51ea3fb20ebd8819e6412093
train_22865
2026-01-01T00:00:00
Governance, provenance, and licensing for code data
agent_loop
intermediate
Task: agent_loop Topic: Governance, provenance, and licensing for code data 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. Loop: Plan → Edit → Test → Reflect → Human gate
[]
{ "target_language": "SQL", "developer_needs": [ "auditability", "reproducibility", "ci_integration", "tooling" ], "moe_experts": [ "performance_expert", "agentic_systems_expert" ], "governance": { "audit_required": true, "tests_required": true } }
2b8c535d2afecf6071dd5e8ee9aa6ec63deeb29f
train_22866
2026-01-01T00:00:00
Governance, provenance, and licensing for code data
design
advanced
Task: design Topic: Governance, provenance, and licensing for code data 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. 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", "auditability", "security_gates" ], "moe_experts": [ "agentic_systems_expert", "governance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
a32a26cd5d71fa7421068aa662fbfdcb6dd5ef88
train_22867
2026-01-01T00:00:00
Reasoning-first coding models and tunable deliberation
review
advanced
Task: review Topic: Reasoning-first coding models and tunable deliberation 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": [ "tooling", "documentation", "ci_integration", "auditability" ], "moe_experts": [ "data_curation_expert", "security_expert" ], "governance": { "audit_required": true, "tests_required": true } }
76a46615dcc12079f79e617b651d1a619c0611a4
train_22868
2026-01-01T00:00:00
Reasoning-first coding models and tunable deliberation
patch_diff
expert
Task: patch_diff Topic: Reasoning-first coding models and tunable deliberation Difficulty: expert Target language: TypeScript 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": "TypeScript", "developer_needs": [ "tests_are_truth", "ci_integration", "governance", "documentation" ], "moe_experts": [ "evaluation_expert", "coding_expert" ], "governance": { "audit_required": true, "tests_required": true } }
e0a4b725e4fc90b38061e77983957f58f4f9eca2
train_22869
2026-01-01T00:00:00
Model merging, distillation, and continued pretraining
patch_diff
intermediate
Task: patch_diff 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. 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", "tooling", "ci_integration", "reproducibility" ], "moe_experts": [ "governance_expert", "evaluation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
460df273a8cd7744bc56dd28c8237deb7106d7ff
train_22870
2026-01-01T00:00:00
SWE-bench style real-repo evaluation
review
expert
Task: review Topic: SWE-bench style real-repo evaluation Difficulty: expert 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. 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": [ "ci_integration", "documentation", "reproducibility", "security_gates" ], "moe_experts": [ "governance_expert", "evaluation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
354b33c143e4ac2a00b7cbecb7d8b5d9bf959f7f
train_22871
2026-01-01T00:00:00
Tool calling, sandboxes, and CI integration
compare
expert
Task: compare Topic: Tool calling, sandboxes, and CI integration Difficulty: expert 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. Compare: capability, cost, latency, reliability
[]
{ "target_language": "TypeScript", "developer_needs": [ "tooling", "tests_are_truth", "auditability", "repo_scale_reasoning" ], "moe_experts": [ "coding_expert", "evaluation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
f21d936857101d14163880b600a0415a4fa77797
train_22872
2026-01-01T00:00:00
Mixture-of-Experts (MoE) for code
review
expert
Task: review Topic: Mixture-of-Experts (MoE) for code 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. Review: correctness, security, performance, governance
[]
{ "target_language": "Go", "developer_needs": [ "security_gates", "tooling", "auditability", "cost_latency_tradeoffs" ], "moe_experts": [ "evaluation_expert", "data_curation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
39ffb8c98aa2374368f300b1c73d6ee9d6c8bacb
train_22873
2026-01-01T00:00:00
Dataset curation pipelines (filter, dedupe, quality)
design
expert
Task: design Topic: Dataset curation pipelines (filter, dedupe, quality) 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
[]
{ "target_language": "TypeScript", "developer_needs": [ "documentation", "tooling", "governance", "tests_are_truth" ], "moe_experts": [ "governance_expert", "security_expert" ], "governance": { "audit_required": true, "tests_required": true } }
ee3be44609c1be7e06fa1ba006ebff359d1bf56e
train_22874
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: 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
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "C#", "developer_needs": [ "documentation", "tooling", "ci_integration", "evaluation_metrics" ], "moe_experts": [ "security_expert", "coding_expert" ], "governance": { "audit_required": true, "tests_required": true } }
0f53be0b5288d2a747575a5b78eafa08dcd53a47
train_22875
2026-01-01T00:00:00
Self-improving agents and feedback loops
data_pipeline
intermediate
Task: data_pipeline Topic: Self-improving agents and feedback loops Difficulty: intermediate 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. Pipeline: Ingest → Normalize → Filter → Dedupe → Quality → Mix → Audit
[]
{ "target_language": "Go", "developer_needs": [ "documentation", "evaluation_metrics", "ci_integration", "auditability" ], "moe_experts": [ "evaluation_expert", "agentic_systems_expert" ], "governance": { "audit_required": true, "tests_required": true } }
918d0744b72cdad74711c1391fef3aa2704f18e9
train_22876
2026-01-01T00:00:00
Reasoning-first coding models and tunable deliberation
explain
intermediate
Task: explain Topic: Reasoning-first coding models and tunable deliberation Difficulty: intermediate 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. 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": [ "governance", "auditability", "cost_latency_tradeoffs", "evaluation_metrics" ], "moe_experts": [ "agentic_systems_expert", "security_expert" ], "governance": { "audit_required": true, "tests_required": true } }
cf6003dff70abf96796778c9a6724cf3a6e4baa7
train_22877
2026-01-01T00:00:00
Tool calling, sandboxes, and CI integration
review
intermediate
Task: review Topic: Tool calling, sandboxes, and CI integration Difficulty: intermediate 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. Review: correctness, security, performance, governance
[]
{ "target_language": "Python", "developer_needs": [ "security_gates", "tooling", "evaluation_metrics", "documentation" ], "moe_experts": [ "governance_expert", "evaluation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
54de4d6d318669347357ff35cc9bb08e6b74df20
train_22878
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: 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": [ "repo_scale_reasoning", "security_gates", "documentation", "tooling" ], "moe_experts": [ "coding_expert", "performance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
27beaaeeab855d2314c36ff885a62c77f792b645
train_22879
2026-01-01T00:00:00
Mixture-of-Experts (MoE) for code
design
advanced
Task: design Topic: Mixture-of-Experts (MoE) for code Difficulty: advanced Target language: Bash 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": "Bash", "developer_needs": [ "documentation", "repo_scale_reasoning", "tests_are_truth", "security_gates" ], "moe_experts": [ "evaluation_expert", "data_curation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
f566b71874b6ba09033e7061cb5de2a1ce2d00ed
train_22880
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: 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. Pipeline: Ingest → Normalize → Filter → Dedupe → Quality → Mix → Audit
[]
{ "target_language": "Rust", "developer_needs": [ "governance", "cost_latency_tradeoffs", "security_gates", "evaluation_metrics" ], "moe_experts": [ "data_curation_expert", "security_expert" ], "governance": { "audit_required": true, "tests_required": true } }
6ce9f87a599d6e5b34f1ad92df8fa76c35d12338
train_22881
2026-01-01T00:00:00
Latency, cost, and reliability optimization
failure_analysis
expert
Task: failure_analysis Topic: Latency, cost, and reliability optimization Difficulty: expert 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. Failure: - Initial patch broke edge case Reflection: - Missing zero-input guard Correction: - Add explicit validation + test
[]
{ "target_language": "C#", "developer_needs": [ "reproducibility", "repo_scale_reasoning", "cost_latency_tradeoffs", "security_gates" ], "moe_experts": [ "agentic_systems_expert", "evaluation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
c813e9869ee4d546939b6bc59e7434ccea18ac89
train_22882
2026-01-01T00:00:00
Extended context and repo-scale understanding
compare
intermediate
Task: compare Topic: Extended context and repo-scale understanding 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. Compare: capability, cost, latency, reliability
[]
{ "target_language": "SQL", "developer_needs": [ "tooling", "cost_latency_tradeoffs", "reproducibility", "evaluation_metrics" ], "moe_experts": [ "evaluation_expert", "data_curation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
974eb9bb0e474efa52d7ba6bd689913fe242015b
train_22883
2026-01-01T00:00:00
Model merging, distillation, and continued pretraining
data_pipeline
advanced
Task: data_pipeline Topic: Model merging, distillation, and continued pretraining Difficulty: advanced Target language: Bash 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": "Bash", "developer_needs": [ "governance", "tooling", "auditability", "ci_integration" ], "moe_experts": [ "evaluation_expert", "governance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
127854640f009b147b84edbf373a38a71b5b2476
train_22884
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: 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. Design with risks, metrics, acceptance criteria
[]
{ "target_language": "Go", "developer_needs": [ "reproducibility", "documentation", "security_gates", "tests_are_truth" ], "moe_experts": [ "performance_expert", "evaluation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
ef6fcbeacdf676c062849eb0eca205e07ee82542
train_22885
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: Bash 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
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "Bash", "developer_needs": [ "reproducibility", "security_gates", "tests_are_truth", "tooling" ], "moe_experts": [ "governance_expert", "agentic_systems_expert" ], "governance": { "audit_required": true, "tests_required": true } }
e538feef7b8e008605d2d66da57a8c9334bc933d
train_22886
2026-01-01T00:00:00
Mixture-of-Experts (MoE) for code
eval
intermediate
Task: eval Topic: Mixture-of-Experts (MoE) for code 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. 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": "C#", "developer_needs": [ "governance", "security_gates", "documentation", "auditability" ], "moe_experts": [ "performance_expert", "security_expert" ], "governance": { "audit_required": true, "tests_required": true } }
19af14e3490976295aea01c4f7b1aec3a29fa722
train_22887
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: SQL 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": "SQL", "developer_needs": [ "cost_latency_tradeoffs", "security_gates", "auditability", "repo_scale_reasoning" ], "moe_experts": [ "security_expert", "governance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
026210b43ad7cd34b632692c3ccdc7af69215201
train_22888
2026-01-01T00:00:00
Latency, cost, and reliability optimization
design
intermediate
Task: design Topic: Latency, cost, and reliability optimization Difficulty: intermediate 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": [ "security_gates", "cost_latency_tradeoffs", "evaluation_metrics", "documentation" ], "moe_experts": [ "coding_expert", "data_curation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
305ab2356b5ecfaf2a798c8107c0dd075366bf08
train_22889
2026-01-01T00:00:00
SWE-bench style real-repo evaluation
agent_loop
intermediate
Task: agent_loop Topic: SWE-bench style real-repo evaluation 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": [ "repo_scale_reasoning", "documentation", "evaluation_metrics", "reproducibility" ], "moe_experts": [ "governance_expert", "agentic_systems_expert" ], "governance": { "audit_required": true, "tests_required": true } }
54e134a00b8de70d94962bbeeef60003e717d248
train_22890
2026-01-01T00:00:00
Mixture-of-Experts (MoE) for code
explain
intermediate
Task: explain Topic: Mixture-of-Experts (MoE) for code 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
[ "Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions" ]
{ "target_language": "Rust", "developer_needs": [ "documentation", "security_gates", "auditability", "tooling" ], "moe_experts": [ "security_expert", "performance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
10c07a9440493a2db532665880287c523cc9b313
train_22891
2026-01-01T00:00:00
Model merging, distillation, and continued pretraining
failure_analysis
intermediate
Task: failure_analysis Topic: Model merging, distillation, and continued pretraining 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. 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": "SQL", "developer_needs": [ "auditability", "repo_scale_reasoning", "reproducibility", "documentation" ], "moe_experts": [ "coding_expert", "data_curation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
fdfbf2bb23f99fa0c009ff06441a9b6077b1dff8
train_22892
2026-01-01T00:00:00
Multimodal dev workflows (docs, diagrams, traces)
compare
intermediate
Task: compare Topic: Multimodal dev workflows (docs, diagrams, traces) Difficulty: intermediate 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": [ "tests_are_truth", "cost_latency_tradeoffs", "auditability", "evaluation_metrics" ], "moe_experts": [ "coding_expert", "security_expert" ], "governance": { "audit_required": true, "tests_required": true } }
7b917e4c0e2ded63401d46eb7432c2c5ea58b49e
train_22893
2026-01-01T00:00:00
Governance, provenance, and licensing for code data
data_pipeline
intermediate
Task: data_pipeline Topic: Governance, provenance, and licensing for code data 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. 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": "Python", "developer_needs": [ "tooling", "tests_are_truth", "governance", "repo_scale_reasoning" ], "moe_experts": [ "performance_expert", "agentic_systems_expert" ], "governance": { "audit_required": true, "tests_required": true } }
4ed0f2c3c68ef84f2a6df5044dc6f2683be482f4
train_22894
2026-01-01T00:00:00
Governance, provenance, and licensing for code data
eval
expert
Task: eval Topic: Governance, provenance, and licensing for code data Difficulty: expert 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. 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": "Rust", "developer_needs": [ "governance", "cost_latency_tradeoffs", "evaluation_metrics", "documentation" ], "moe_experts": [ "performance_expert", "agentic_systems_expert" ], "governance": { "audit_required": true, "tests_required": true } }
e48abd6c5fb4bb0c1d7e98efa041c22ac0ec01a8
train_22895
2026-01-01T00:00:00
Latency, cost, and reliability optimization
patch_diff
advanced
Task: patch_diff Topic: Latency, cost, and reliability optimization Difficulty: advanced 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. 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": "Bash", "developer_needs": [ "repo_scale_reasoning", "documentation", "ci_integration", "reproducibility" ], "moe_experts": [ "security_expert", "coding_expert" ], "governance": { "audit_required": true, "tests_required": true } }
44e21cdad21ea2313df2078dacf18e1eef9c7214
train_22896
2026-01-01T00:00:00
Self-improving agents and feedback loops
patch_diff
expert
Task: patch_diff 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. 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": [ "tests_are_truth", "tooling", "auditability", "governance" ], "moe_experts": [ "security_expert", "governance_expert" ], "governance": { "audit_required": true, "tests_required": true } }
1f25f15f3e4b0232509f8c6957ad96fcf32e4a96
train_22897
2026-01-01T00:00:00
Dataset curation pipelines (filter, dedupe, quality)
review
expert
Task: review Topic: Dataset curation pipelines (filter, dedupe, quality) Difficulty: expert 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. 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": [ "auditability", "cost_latency_tradeoffs", "documentation", "tests_are_truth" ], "moe_experts": [ "performance_expert", "data_curation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
cf89c0ca4101522974a13ce60c4a8132bb0ff251
train_22898
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: 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": [ "repo_scale_reasoning", "cost_latency_tradeoffs", "tests_are_truth", "documentation" ], "moe_experts": [ "governance_expert", "security_expert" ], "governance": { "audit_required": true, "tests_required": true } }
a4dbf7b19c14c99197d85c7fd7870968237afddc
train_22899
2026-01-01T00:00:00
Mixture-of-Experts (MoE) for code
review
intermediate
Task: review Topic: Mixture-of-Experts (MoE) for code 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. Review: correctness, security, performance, governance
[]
{ "target_language": "Python", "developer_needs": [ "auditability", "reproducibility", "repo_scale_reasoning", "ci_integration" ], "moe_experts": [ "agentic_systems_expert", "evaluation_expert" ], "governance": { "audit_required": true, "tests_required": true } }
b8764705f751db6c529c210f10fca2da487e4c93