Terryc21

Terryc21 / skill-reviewer

Public

Candid reviews of Claude Code skills. File:line citations. Ranked actions. No filler.

15
4
89% credibility
Found May 18, 2026 at 16 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
AI Summary

skill-reviewer is a reviewer tool for Claude Code that gives honest, structured feedback on AI assistant skills. It reads through skill files, identifies both strengths and weaknesses, and produces detailed reports with specific findings organized by severity. The tool supports quick summaries, comprehensive audits, and focused reviews on specific aspects like safety, architecture, or test coverage. It also offers a 'second opinion' mode that independently challenges its top conclusions before finalizing the report.

How It Works

1
💡 You just finished building something

You've created a skill (a helpful automation tool) and want to know if it's actually good before sharing it with others.

2
📦 You add the reviewer to your assistant

You install skill-reviewer as an extension to your AI coding assistant with two simple commands, and it becomes available to help you.

3
🎯 You ask it to review your work

You point the reviewer at your skill folder and ask it to take a look, specifying if you want a full audit or a quick summary.

4
Choose your review depth
⚡
Quick summary

Get the headline verdict, a few key strengths, and top issues in under a minute.

📋
Full audit

Get everything: detailed strengths, issue cards with specific suggestions, and patterns across your work.

5
📄 You receive your structured report

Within a few minutes, you get a complete report with a clear verdict, specific findings backed by actual examples from your files, and organized by what matters most.

6
✨ You act on the feedback

The review cards are sorted so you can tackle the easy wins first, and each finding tells you exactly what to improve and why it matters.

🚀 Your skill is ready to ship

You fixed the issues, kept what works, and now you have honest confidence that your skill is solid before publishing it.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 16 to 15 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is skill-reviewer?

This is a Claude Code skill that reviews other Claude Code skills. You point it at a skill directory and get back structured, candid feedback with specific line citations, severity tags, and a clear separation of strengths from weaknesses. The output includes a TL;DR, 5-10 strengths to keep, finding cards sorted by severity (each with Why/Fix/citation/effort/quick-win tags), and a Patterns section synthesizing what crossed cards. There's also a summary mode for quicker audits and lenses for focused reviews on specific concerns like safety, architecture, or test coverage.

Why is it gaining traction?

The hook is honesty. It refuses to hedge—if something is fragile, it says so directly. The lens system lets you audit just what you care about without the full review overhead. Second-opinion mode adds a challenger that reads the cited files itself, not just the draft quotes, which actually challenges your top conclusions before you publish. The whole thing is designed for the moment when you've stared at your own work too long and can't see what's missing.

Who should use this?

Claude Code skill authors who want honest feedback before publishing on GitHub or a marketplace. Developers auditing someone else's skill without inventing a review structure from scratch. Teams merging substantial PRs to a skill who need a structured pre-merge audit. Anyone who suspects their own review is too kind (or too harsh) and wants a second opinion.

Verdict

With a credibility score of 0.9% and only 15 stars, this is a niche tool at an early maturity stage. That said, the documentation is thorough, the self-review mechanism keeps quality high, and it fills a real gap for the skill-author community. If you write Claude Code skills, it's worth a try. Just don't expect it to review non-Anthropic skills or handle high-volume workflows yet.

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.