Reviews uncommitted code changes for design, bugs, and requirements before final submission.

Install

mkdir -p .claude/skills/review-matrixages && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/9502" && unzip -o skill.zip -d .claude/skills/review-matrixages && rm skill.zip

Installs to .claude/skills/review-matrixages

Activation

This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.

Review uncommitted code changes to find unreasonable design, unfinished requirements, potential bugs, missed reuse opportunities, and optimization opportunities. Use when the user asks for a review of pending, uncommitted, or unpublished code changes and expects both findings and concrete improvements.
303 chars✓ has a “when” triggerlonger than Claude Code's old 250-char listing cap (fine on current versions)
Beginner

Key capabilities

  • Identify fragile design choices
  • Detect unfinished requirements
  • Find potential bugs and regressions
  • Locate duplicated logic
  • Verify missing test coverage

How it works

The tool inspects the working tree using git status and diffs to identify issues before providing actionable feedback or applying fixes.

Inputs & outputs

You give it
uncommitted code changes
You get back
list of findings and suggested fixes

When to use review

  • Reviewing uncommitted changes for potential bugs
  • Identifying duplicated logic in new code
  • Ensuring code requirements are fully implemented
  • Checking for missing test coverage

About this skill

Review

Inspect the current working tree before making assumptions. Start with git status --short, then read the relevant diff.

Prioritize findings over summaries. Report the highest-impact issues first with precise file references and concrete impact.

Check these areas on every run:

  • Find unreasonable or fragile design choices.
  • Find unfinished requirements or behavior gaps.
  • Find potential bugs, regressions, and edge-case failures.
  • Find duplicated logic or places where existing code should have been reused.
  • Find missing validation, verification, or tests.

When the user expects action instead of commentary, fix the issues you identify and re-check the result.

When reviewing, keep the response structure tight:

  • List findings first, ordered by severity.
  • Add open questions or assumptions only if they materially affect correctness.
  • Add a short change summary only after the findings or fixes.

When no issues are found, say so explicitly and still mention residual risks or missing verification.

When not to use it

  • When code changes are already committed or published

Limitations

  • Requires uncommitted changes to be present in the working tree

How it compares

It automates the review of pending changes by prioritizing high-impact issues and providing concrete fixes rather than just general commentary.

Compared to similar skills

review side by side with the closest alternatives in the catalog.

SkillInstallsUpdatedSafetyDifficulty
review (this skill)02moNo flagsBeginner
python-testing-patterns772moReviewIntermediate
fix-bug117moReviewIntermediate
test-fixing19moReviewIntermediate

Try saying

Example prompts that trigger this skill in your AI assistant.

Search skills

Search the agent skills registry