Tune review signal with current learnings, detected guideline files, path instructions, and measured feedback. Use when this operator task needs a current, evidence-backed CodeRabbit workflow. Trigger with "tune CodeRabbit reviews".
Improve relevance using supported knowledge layers. Avoid invented configuration keys and judge changes against labeled real findings.
references/official-docs.md and re-check any time-sensitive contract before execution.CLAUDE.md and AGENTS.md.Treat Git-provider sessions, CodeRabbit web sessions, CLI credentials, and CodeRabbit API keys as separate credentials. Use only an already-approved session or secret-manager reference, never print a secret, and do not place credentials in .coderabbit.yaml, source files, logs, or deliverables.
Sample accepted, rejected, and unresolved findings.
Separate organization guidance, repository rules, path instructions, and temporary feedback.
Draft one supported measurable tuning change and validate resolved config.
Pilot comparable PRs and retain only improvements that preserve risky-path coverage.
Require owner approval before changing shared learnings, global overrides, or exclusions. Keep analysis and drafts local until approval is explicit, and record who approved the action and its scope.
A labeled sample, hypothesis, supported change, comparison metrics, and rollback decision. Include source dates, unknowns, and the exact boundary between observed fact and recommendation.
| Condition | Response |
|---|---|
| Current contract is unclear or docs disagree | Stop mutation, cite both sources, and request owner resolution. |
| Required access or approval is missing | Produce a draft and evidence plan only. |
| Validation or pilot behavior differs from expectation | Restore the prior state and retain the failed evidence. |
| Output contains secrets or private code | Stop, quarantine the artifact, redact it, and notify the data owner. |
Move a durable convention into AGENTS.md and verify detection.
Replace unsupported custom_patterns with documented path instructions.
references/official-docs.md.88dcf65
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.