Skill Assessor

About

Empowers developers to maintain a lean and efficient Claude Code environment by systematically analyzing the impact of new skills before they are added to global rules. By coordinating specialized context-gathering, code-analyzer, and research agents, it calculates a rigorous value score that balances safety guardrails and usage frequency against potential token overhead. This ensures that only high-value, high-relevance skills are set to auto-invoke, preventing skill bloat while maximizing the model's ability to provide proactive assistance in critical scenarios.

Key Features

• Automated codebase pattern analysis to determine trigger frequency
• 2 GitHub stars
• Quantitative value scoring based on relevance, impact, and token cost
• Prioritized three-tier evaluation criteria: Safety, Frequency, and Convenience
• Structured recommendations for trigger keywords and intent patterns
• Multi-agent orchestration using specialized analysis and research agents

Use Cases

• Optimizing token usage by identifying skills with high noise-to-value ratios
• Ensuring safety-critical guardrail skills are properly configured for consistent enforcement
• Determining if a newly developed custom skill should be globally auto-invoked or manually called