skill-judge

To evaluate AI skills, you can use an evaluation protocol and scoring rubric to score them across dimensions like Knowledge Delta, Mindset, and Anti-Patterns for a total score out of 120.

Benchmarking AI skills involves systematically auditing SKILL.md files against official specifications to compare effectiveness, identify optimization opportunities, and ensure adherence to established patterns.

Yes, you can perform quality audits on collections of AI skills to systematically identify those needing optimization by evaluating their adherence to defined specifications and established patterns.

A skill evaluation rubric scores SKILL.md files by assessing 8 distinct dimensions, including Knowledge Delta and Anti-Patterns, resulting in a total effectiveness score out of 120 points.

You need a quality audit when you have developed multiple AI skills and want to ensure they meet a high standard before deployment, identifying weaknesses to guide improvements and optimization.

No specific dependencies are required for skill evaluation. The process operates directly on SKILL.md files, applying an established evaluation protocol and scoring rubric to measure quality and effectiveness.