Audits Claude Code skills against design best practices for token efficiency, structure, and modularization.
The auditing skill provides a robust framework for evaluating Claude Code skills to ensure they meet high-performance standards. It systematically scans SKILL.md files and reference directories to measure token budgets, validate workflow structures, and check for modularization patterns. By identifying over-budget files and common implementation 'gotchas'—such as improper shell metacharacter usage or name collisions—this tool helps developers build lean, efficient, and reliable skills that minimize token costs and maximize reliability within the Claude Code ecosystem.
Key Features
01Standardized audit reporting with pass/fail scoring and actionable recommendations
02Automated token budget analysis to prevent prompt bloat and latency
030 GitHub stars
04Modularization checks to ensure large content blocks are loaded conditionally
05Structural validation of frontmatter, workflow steps, and reporting patterns
06Detection of common authoring pitfalls including syntax and naming errors
Use Cases
01Pre-publication validation for skills intended for the Skillforge marketplace
02Ensuring consistency and quality across a library of custom team skills
03Optimizing complex skill files to reduce token consumption and improve response speed
