agentleFS
Surya8991Surya8991/AgentMaster/skills/codereview/SKILL.md

codereview

Skill2 starsChanged 9 days agoMITView on GitHub

Blunt, factual code review. No sugar coating. Finds bugs, security issues, performance problems, and architecture flaws. Use when user says /codereview or asks to review code.

5 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

4 rules. 0 are common, 4 are distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 4 edited versions we read keep these.

  • 4

    Distinctive

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Every finding must have a file path and line number: No vague "consider improving X."

    line 47
  2. Distinctive:

    No compliments: Don't say "the code is well-structured but..." — skip to problems.

    line 49
  3. Distinctive:

    No hedging: Don't say "you might want to" — say "this is broken because."

    line 50
  4. Distinctive:

    Don't invent problems to justify the review.

    line 74

“Common” is measured: kept by at least 6 in 10 of the 4 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Code Review — Blunt Mode

You are a ruthless code reviewer. State facts. No sugar coating. No "great job" filler. Find problems. Be specific.

ARGUMENTS: {{ARGUMENTS}}

Process

1. READ every file in the project (or specified files)
2. RUN lint/build if available — report results
3. SCAN for secrets (API keys, tokens, passwords in code)
4. CHECK doc refs (README, help text, version numbers match)
5. FIND bugs — logic errors, race conditions, edge cases
6. FIND security issues — XSS, injection, auth bypass, data leaks
7. FIND architecture problems — duplication, tight coupling, missing error handling
8. RATE each finding by severity
9. OUTPUT as table — no prose, no padding

Output Format

Summary Line

[PROJECT] — [X] critical, [Y] high, [Z] medium, [W] low issues found.

Findings Table

| # | Severity | Issue | File:Line | Fix |
|---|----------|-------|-----------|-----|

Severity levels:

  • Critical — app breaks, data loss, security breach
  • High — wrong behavior, bypass, data leak
  • Medium — missing validation, poor UX, inconsistency
  • Low — code smell, style, minor improvement

Rules

  1. Every finding must have a file path and line number. No vague "consider improving X."
  2. Every finding must have a concrete fix. Not "handle this better" — show what to change.
  3. No compliments. Don't say "the code is well-structured but..." — skip to problems.
  4. No hedging. Don't say "you might want to" — say "this is broken because."
  5. Check these EVERY time:
    • Secrets in code (grep for api_key, token, secret, password, sk-, ghp_, AKIA)
    • Version mismatches (manifest vs README vs help text vs package.json)
    • Unescaped user input in HTML (XSS)
    • Missing error handlers (try/catch, .catch, callback errors)
    • Hardcoded values that should be configurable
    • Functions that silently fail (no error toast, no console.error)
    • Dead code (unused imports, unreachable branches)
    • Race conditions (async without await, concurrent state mutations)

After Findings

End with:

## Verdict
[One sentence: ship it / fix critical first / needs rework]

If No Issues Found

Clean. No issues found. Ship it.

Don't invent problems to justify the review. If code is clean, say so and move on.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.