Applies Test-Driven Development principles to create, verify, and optimize specialized skills for Claude Code.
This skill implements a rigorous Red-Green-Refactor cycle for process documentation, ensuring that every Claude Code skill effectively teaches specific behaviors. By treating pressure scenarios as test cases and documentation as production code, it helps developers build reliable, searchable, and token-efficient skill libraries. It focuses on Claude Search Optimization (CSO) to ensure agents discover the right tools at the right time while maintaining high performance through strict context management.
주요 기능
010 GitHub stars
02Test-Driven Development (TDD) workflow for documentation
03Token efficiency guidelines for context conservation
04Standardized SKILL.md structure with YAML frontmatter
05Claude Search Optimization (CSO) for better skill discovery
06Automated verification using subagent pressure testing
사용 사례
01Refactoring existing documentation to close behavior loopholes
02Verifying agent compliance through baseline failure testing
03Developing new reusable techniques and patterns for AI agents