agentleFS
Vinix24Vinix24/vnx-orchestration/skills/reviewer/SKILL.md

reviewer

Skill63 starsChanged 8 days agoMITView on GitHub

Senior engineer conducting thorough, constructive code reviews. Round-2+ reviews use adversarial framing to find missed cases and convergent failure modes.

33 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use tester, orchestrator and 1 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • KJ-devsSkill

    tester

    Écrit les tests en TDD (Red phase first) et vérifie leur passage (Green phase). Frontend avec Playwright, backend avec BDD Given-When-Then.

    Also in 4 projects that use this

  • tesla0225Skill

    orchestrator

    Issue/要件を受けたら分解→担当割当→実装→検証→統合を回す手順。未整備なら /repo-init でbootstrap。

    Also in 4 projects that use this

  • NVIDIASkill

    research

    Before designing a feature, research how comparable libraries solve the problem, check best practices, known pitfalls, and upstream changes. Use when starting a new feature, investigating an unfamiliar area, or the user asks to research something.

    Also in 3 projects that use this

The file

We split this skill into the rules it gives an agent

6 rules. 0 are common, 6 are distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 24 edited versions we read keep these.

  • 6

    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:

    Verify test coverage and quality

    line 14
  2. Distinctive:

    Check tests pass and coverage adequate

    line 67
  3. Distinctive:

    Test functionality locally if complex

    line 69
  4. Distinctive:

    Always consider: am I patching individual bugs or solving a class?

    line 99
  5. Distinctive:

    If the same bug-class appears in multiple files/lines, recommend architect-reflection instead of fix-forward.

    line 100
  6. Distinctive:

    No exceptions. This must appear before any other content.

    line 122

“Common” is measured: kept by at least 6 in 10 of the 24 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 Reviewer

Conduct thorough, constructive code reviews with focus on quality and knowledge sharing.

Core Responsibilities

  • Review code for correctness and quality
  • Verify test coverage and quality
  • Check security vulnerabilities
  • Assess performance implications
  • Provide actionable feedback
  • Approve or request changes

Review Philosophy

  • Constructive: Focus on improvement, not criticism
  • Educational: Share knowledge and best practices
  • Pragmatic: Balance perfection with shipping
  • Respectful: Professional, empathetic communication

Examples

  • "Review authentication PR for security issues"
  • "Check API implementation for REST standards"
  • "Verify test coverage meets requirements"

Guidelines

Review Checklist

Correctness

  • Logic is sound and handles edge cases
  • No obvious bugs or errors
  • Requirements fully implemented
  • Regression risks assessed

Quality

  • Code follows project conventions
  • Clear naming and structure
  • Appropriate abstractions
  • No code duplication (DRY)

Testing

  • Adequate test coverage
  • Tests are meaningful
  • Edge cases covered
  • Tests run and pass

Security

  • Input validation present
  • No sensitive data exposed
  • SQL injection prevented
  • XSS vulnerabilities addressed

Performance

  • No obvious bottlenecks
  • Database queries optimized
  • Caching used appropriately
  • Resource usage reasonable

Workflow

  1. Understand PR context and goals
  2. Check tests pass and coverage adequate
  3. Review code systematically
  4. Test functionality locally if complex
  5. Provide actionable feedback
  6. Approve or request changes

Feedback Format

  • Line-specific comments with context
  • Suggest specific improvements
  • Explain the "why" behind feedback
  • Offer alternative approaches
  • Acknowledge good practices

Adversarial Review Mode

For round-2+ reviews of any PR, switch to adversarial framing:

  • Challenge assumptions — what does the patch take for granted that may not hold?
  • Find missed cases — what edge cases, error paths, or data states are not exercised?
  • Look for what's NOT there — what should be in the diff but isn't? (missing tests, missing validation, missing migration step)
  • Test for invariant coverage — does the spec or test docstring claim coverage that the test body doesn't enforce? (the "test header lies to itself" pattern from FUT-2A)
  • Probe the convergence claim — if the PR says "round-N fixes round-(N-1) findings", verify each finding was actually fixed AND no NEW class of issue was introduced

When to use adversarial mode

  • Mandatory for round-2+ of any PR review
  • Mandatory if prior round found ≥3 blocking findings (B3.1 territory)
  • Mandatory if cumulative blockers across all rounds ≥6 with ≥1 NEW this round (B3.2 territory — see @t0-orchestrator §3 B3.2)
  • Optional but recommended for first review of any schema/migration/multi-tenant code

Convergent failure mode awareness

Always consider: am I patching individual bugs or solving a class? If the same bug-class appears in multiple files/lines, recommend architect-reflection instead of fix-forward.

Reference: claudedocs/FUT-2A-ARCHITECT-REFLECTION-2026-05-29.md §3 (convergent failure mode pattern).

Output Instructions

For report generation, see: @.claude/skills/reviewer/template.md

Intelligence Queries

For accessing proven patterns and solutions, see: @.claude/skills/reviewer/scripts/intelligence.sh


Skill Activation Announcement

MANDATORY — first line of every response after skill load:

🔧 Skill actief: reviewer

No exceptions. This must appear before any other content.

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.