Can it help fix skills that aren't working correctly?

Yes, it includes a dedicated section for identifying anti-patterns and applying improvement patterns to troubleshoot and resolve guidance or trigger issues.

What is the primary goal of the Agent Skill Reviewer?

It provides a systematic framework for auditing and improving AI agent skills to ensure they are effective, maintainable, and compliant with the agent-skills standard.

What dimensions are used to evaluate skill quality?

Skills are evaluated based on four key criteria: Correctness (specification compliance), Clarity (unambiguous instructions), Effectiveness (triggering/guidance), and Maintainability (organization).

How does it improve the code review process for skills?

It provides standardized feedback templates, severity guidelines, and a checklist to ensure reviews are thorough, actionable, and consistent.

Agent Skill Reviewer

Name: Agent Skill Reviewer
Author: bitsydarel

bybitsydarel

0•

Collaboration Tools

Audits and optimizes AI agent skills to ensure high-quality, effective, and compliant task execution across development workflows.

The Agent Skill Reviewer provides a systematic framework for evaluating and optimizing Claude Code skills against rigorous quality standards. It streamlines the review process for pull requests, identifies common anti-patterns such as style-only focus or lack of actionable feedback, and offers proven improvement patterns to enhance instruction clarity and effectiveness. By evaluating skills across the dimensions of correctness, clarity, effectiveness, and maintainability, this skill serves as a comprehensive quality assurance tool for developers looking to build more reliable and predictable AI agents.

Key Features

01Standardized feedback templates with severity guidelines for code reviews

02Anti-pattern identification with specific detection methods and fixes

030 GitHub stars

04Systematic evaluation across correctness, clarity, effectiveness, and maintainability

05Structural validation for directory and file organization compliance

06Reference library of improvement patterns for troubleshooting underperforming skills

Use Cases

01Troubleshooting and refining skills that fail to trigger or provide poor guidance

02Conducting detailed peer reviews on new AI skill pull requests

03Auditing existing skill libraries to ensure consistent standards and maintainability

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add bitsydarel/bdskills bd-skill-reviewing

For use in Claude.ai and ChatGPT

Download Skill