Skill Evaluator
Evaluates agent skills against Anthropic's official best practices to ensure optimal structure, naming, and content organization.
The source repository doesn't declare a license. Check its terms before reusing the code.
Key features
- Standardized reporting with scores and prioritized recommendations
- Weighted scoring rubric for naming, description quality, and content
- Automated validation of SKILL.md structure and YAML frontmatter
- Guidance on calibrating degrees of freedom for specific task types
- Detection of common anti-patterns like nested references and backslashes
Use cases
- Identifying content bloat and vague activation triggers in agent instructions
- Auditing a new skill before publishing it to a directory or repository
- Reviewing existing skills for compliance with official Anthropic guidelines
FAQ
How does this skill improve my AI development workflow?
It automates the validation of skill metadata and identifies common anti-patterns like nested references or overly verbose instructions. This ensures Claude can more accurately interpret and execute your custom skills.
What specific criteria does the Skill Evaluator score?
It provides a weighted scoring rubric across six dimensions: Naming (10%), Description (20%), Content Quality (30%), Structure & Organization (25%), Degrees of Freedom (10%), and an Anti-Pattern Check (5%).
When should I use this skill?
Use this skill during the development or review phase of any new Claude capability. It is specifically designed to be used when you need to evaluate, assess, or audit a skill's quality before deployment.
What does the Skill Evaluator do?
The Skill Evaluator audits Claude Code skills by analyzing their SKILL.md structure, naming conventions, and content organization against Anthropic's official best practices to ensure peak performance.
Does it provide actionable feedback?
Yes. Upon evaluation, the skill generates a standardized report that includes an overall score, a breakdown of dimensions, and a prioritized list of recommendations to improve the skill's effectiveness.
Related skills
OpenCode Skill Tester
Verifies and validates the full suite of OpenCode Agent Skills plugin tools, scripts, and file access APIs.
Security Testing27012 ptsTest Skill
Validates the execution of PostToolUse hooks to ensure proper skill invocation and system integration.
Security Testing312 ptsBiomeJS Configuration
Configures BiomeJS for code formatting and linting using best practices for monorepos, UI projects, and Git hooks.
Security Testing5812 pts