Goal
Produce a structured maintenance report that separates urgent repository risks from polish, identifies gaps in onboarding and automation, and recommends the next practical maintainer actions.
When to use
- The user asks whether a repository is healthy, maintainable, or ready to publish.
- A repo needs a maintainer audit across docs, tests, CI, releases, and security hygiene.
- The user wants a prioritized cleanup plan without immediate code edits.
When not to use
- The user asks for a focused PR review; use
pr-review. - The task is only documentation clarity; use
docs-audit. - The user already asked for implementation instead of audit.
Inputs to inspect
- README, docs, examples, changelog, license, security policy, and contribution guide.
- Package files, lockfiles, test config, format/lint config, and validation scripts.
- CI workflows, issue templates, release automation, and dependency update config.
- Current
git status, recent commits, and validation output when available.
Review rubric
Score docs, build/test, CI, release readiness, dependency hygiene, security baseline, issue/PR operations, and agent context from 0 to 5. Read references/audit-rubric.md for scoring detail and references/report-template.md for the report shape.
Workflow
- Inspect top-level files and repository structure.
- Identify the main stack and validation commands.
- Review docs against actual commands and files.
- Check CI, release, security, and contribution surfaces.
- Run safe validation commands when the user expects implementation-level evidence.
- Produce a scored report with urgent issues, quick wins, backlog items, and the next action.
Safety rules
- Do not change repository files unless the user explicitly asks for implementation.
- Do not print secrets or sensitive config values in the audit.
- Mark live repository settings as unverified unless they were checked directly.
References
Read only when needed:
references/audit-rubric.mdreferences/report-template.md
Scripts
No bundled scripts.
Output format
Return:
- Summary
- Scores table
- Highest-risk issues
- Quick wins
- Suggested issue backlog
- Recommended next action
Failure modes
- If validation cannot run, explain the missing tool, dependency, or environment.
- If the repo lacks clear goals, state the inferred goal and mark it as an assumption.
- If live settings are required, say which settings need verification.
Completion criteria
- The report is grounded in file or command evidence.
- Risks are prioritized by maintainer impact.
- The next action is concrete and small enough to assign.