What problem does it solve?
Manually auditing a library of Claude skills and slash commands for quality, overlap, staleness, and alignment with best practices is time-consuming and error-prone, especially when managing skills across global and project-level directories.
Core Features & Use Cases
- Dual audit modes: Quick Scan for recently changed skills (5–10 minutes) and Full Stocktake for comprehensive reviews (20–30 minutes), with automatic result caching for fast repeat scans.
- Holistic quality evaluation: Uses AI judgment to assess each skill’s actionability, scope fit, uniqueness, and currency against a standardized checklist, rather than relying on rigid numeric rubrics.
- Usage-aware prioritization: Factors in how frequently skills are used over 7 and 30 day windows to help prioritize maintenance tasks.
Use case: For teams maintaining dozens of custom Claude skills, this skill automates the tedious work of identifying stale, redundant, or low-value skills to keep the skill library lean and effective.
Quick Start
Invoke the skill-stocktake skill from your project root to audit all global and project-level Claude skills for quality, with automatic caching for fast repeat scans of recently changed skills.
