Standardizes the creation and verification of Claude Code skills using a Test-Driven Development approach.
This skill provides a rigorous framework for developing specialized AI capabilities through Test-Driven Development (TDD) applied to documentation. It guides users through the RED-GREEN-REFACTOR cycle—creating pressure scenarios to baseline AI behavior, writing precise documentation to address specific failures, and optimizing for Claude Search Optimization (CSO) to ensure maximum discovery and token efficiency. By treating documentation like production code, it ensures that every skill added to your directory has a measurable impact on AI performance and reliability.
主な機能
01Cross-referencing logic for complex skill hierarchies
02Token-efficient documentation guidelines to save context
03Standardized SKILL.md frontmatter and structure
042 GitHub stars
05TDD-based documentation workflow for AI agents
06Claude Search Optimization (CSO) triggers and keyword patterns
ユースケース
01Developing reusable domain-specific patterns for Claude Code
02Standardizing process documentation across large development teams
03Refining agent-specific skills to close logic and compliance loopholes
