What does the Reviewing Skills tool check?

It validates the YAML frontmatter, body structure, length (aiming for under 500 lines), terminology consistency, and looks for common anti-patterns in your SKILL.md files.

Can it fix my skills automatically?

Yes, after generating a detailed review report, the skill provides an interactive workflow to apply recommended fixes step-by-step, prioritizing critical issues.

How do I trigger a skill review?

You can trigger it by asking Claude to 'review this skill', 'check skill quality', or 'validate SKILL.md' while in the relevant project directory.

Does it support multiple skills in one project?

Yes, it can search for SKILL.md files in the current directory and allows you to select which specific skill to analyze if multiple are detected.

Why is the 500-line limit important for skills?

Keeping skills concise ensures Claude can process instructions more accurately, maintains better focus, and reduces the risk of context window confusion.

Claude Skill Reviewer

Name: Claude Skill Reviewer
Author: j5ik2o

byj5ik2o

0•

开发者工具

Audits and optimizes Claude Code skills based on official best practices to ensure high-quality AI behaviors.

The Reviewing Skills tool provides a comprehensive auditing framework for Claude Code SKILL.md files, ensuring they adhere to official architectural standards and performance benchmarks. It automatically validates YAML frontmatter, analyzes structural clarity, checks for anti-patterns like magic constants or excessive length, and provides an actionable report categorized by severity. Whether you are building a new skill or refining an existing one, this tool streamlines the quality assurance process to ensure your AI agents perform reliably and efficiently through interactive improvement workflows.

主要功能

010 GitHub stars

02Automated SKILL.md validation against official best practices

03Detection of common anti-patterns like Windows-style paths and magic constants

04Interactive improvement workflow to apply recommended fixes automatically

05Verification of YAML metadata for better discovery and trigger accuracy

06Detailed severity-based reporting for Critical, Warning, and Info issues

使用场景

01Optimizing existing skills that have become too complex or exceed the 500-line recommendation

02Standardizing skill documentation and workflow design across a development team

03Validating a new Claude Code skill before deployment to ensure it meets quality standards

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add j5ik2o/cqrs-es-spec-kit-js reviewing-skills

For use in Claude.ai and ChatGPT

Download Skill