What problem does it solve?
Teams using multiple AI coding agents (Claude, Cursor, Kiro, Copilot) often cannot tell which installed skills deliver value, which are failing, and which are unused, leading to wasted AI spend and reduced agent performance.
Core Features & Use Cases
- Per-Skill KPI Tracking: Aggregates run logs and user feedback to calculate run count, success rate, average duration, agent usage breakdown, and inefficiency scores for every installed skill.
- Clear Skill Ratings: Classifies each skill as Active, Needs attention, Low usage, or Unused based on consistent, transparent thresholds.
- Data-Driven Recommendations: Suggests skills to add that match your project's tech stack, and flags unused or poorly performing skills for potential removal with context on why.
- Use Case: A development team using Claude Code and Cursor can run this skill to quickly audit their installed skills, cut waste from unused tools, and add missing skills that match their project's tech stack.
Quick Start
Use the skill-usage-insights skill to generate a usage and KPI report for all installed skills in this project, with recommendations on which to add or remove.
