agentsclimarketplace

Buap repo audit

Skill codysumpter-cloud/buddy-universal-agent-profile/plugins/buap/skills/buap-repo-audit

Audit a repository and produce verified findings only, never trusting README/doc claims without implementation evidence. Use when asked to audit, assess, review the state of, or "tell me what actually works in" a repo or codebase, or to compare docs against source. Follows the BUAP repo-audit runbook and labels every finding.From its SKILL.md

Install
npx -y skills add codysumpter-cloud/buddy-universal-agent-profile --skill buap-repo-audit

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 2 stars2 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

1.8 KB, as published. Nobody here has run it

BUAP repo audit

Produce verified findings only. Do not trust README or doc claims without implementation evidence. This is the BUAP runbooks/repo-audit.md procedure.

Steps

  1. Read repo-local instructions first: AGENTS.md, CLAUDE.md, CODEX.md, .cursor/rules, README. A repo's own contract outranks generic assumptions.
  2. Map the repository structure (tree, entrypoints, packages).
  3. Identify build / test / package managers actually in use.
  4. Inspect the implementation files related to the user's question — read the code, not just the docs.
  5. Compare docs against source. Flag every place they disagree.
  6. Label each finding:
    • Verified — checked directly here (cite file:line, command output).
    • Source-backed — supported by cited source material, not run.
    • Unverified — plausible but not checked.
    • Blocked — missing tool, permission, file, or credential.
  7. Recommend the single smallest next action.

Validation

  • Every finding cites files, lines, PRs, commits, or command output.
  • State explicitly which checks were NOT run.

Do not

  • Claim functionality exists from docs alone.
  • Rewrite or "improve" architecture during an audit — observe, don't change.
  • Mix assumptions into the verified-findings list.

For non-trivial audits, delegate the inspection sweep to the lil-buddy subagent and review its report before answering.

Keep looking

Skills are one crate of 609,689. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.