[crepuscular-drizzle]

Summary

• Add an AIME adapter README covering the pinned source, runtime dependencies, task splits, scoring, and validation commands.
• Update dataset onboarding to point readers to public adapter documentation. Retire the AIME-specific review blocker and preserve the original review and diagnostic notes in Git history.
• Record the 60-task official evaluation pool and the optional 37/23 hash split. The earlier diagnostic was unmeasured, and no side-by-side benchmark comparison with Harbor is on record.

Validation

• python3 tools/check_dataset_blockers.py and python3 tools/datasets_doc.py --check passed. All 11 pinned historical run records remain intact.
• 56 focused registry, evidence-retention, and AIME scorer tests passed.
• 12 AIME QA contract tests passed with synthetic data. Synthetic split checks passed, and sandboxed oracle and no-op controls scored 1 and 0.
• The dataset lint skipped the unprovisioned AIME release. No model evaluation was run.

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has insufficient remaining quota. Try requesting a review at a lower effort level.

An organization owner can choose the repository organization as the billing source for reviews requested by members with a paid Copilot license. Learn about Copilot code review billing. Review the Copilot code review billing settings.