Monitors and analyzes Claude Code skill usage to optimize token consumption and workflow efficiency.
The Skill Performance Profiler is a comprehensive diagnostic tool designed to track how Claude Code skills are utilized across conversations. It extracts deep metrics such as invocation frequency and estimated token consumption, categorizing skills from 'lightweight' to 'very heavy' based on their resource footprint. By identifying co-occurrence patterns and suggesting consolidation opportunities, this skill helps users streamline their AI environment, reduce token overhead, and maintain a high-performance workspace through detailed markdown reports, CSV exports, and visual trend analysis.
주요 기능
01Automated reporting in Markdown and CSV formats
02Automatic categorization of skills based on resource intensity
030 GitHub stars
04Granular token consumption tracking for every skill invocation
05Co-occurrence analysis to identify skills that should be merged
06Time-based trend analysis of skill performance metrics
사용 사례
01Auditing skill usage to remove redundant or underutilized tools
02Finding consolidation opportunities where skills are frequently used together
03Identifying which skills are driving up API costs via high token usage
