Standardizes the creation and optimization of Claude skills using a test-driven documentation framework.
This skill provides a rigorous framework for documenting AI capabilities and patterns by applying Test-Driven Development (TDD) principles to process documentation. It ensures skills are discoverable and effective by teaching users how to write pressure scenarios, observe baseline failures, and refactor instructions to close logic loopholes. Key focus areas include Claude Search Optimization (CSO), token efficiency for long-term context savings, and strategic description writing to prevent agents from skipping complex workflows.
主な機能
010 GitHub stars
02Test-Driven Development (TDD) mapping for process documentation
03Claude Search Optimization (CSO) for improved skill discovery
04Token-efficient documentation patterns and cross-referencing
05Graphviz flowchart conventions for non-obvious decision points
06Standardized SKILL.md directory structure and YAML frontmatter
ユースケース
01Developing reusable domain-specific capabilities for Claude Code
02Optimizing agent performance by preventing documentation shortcuts
03Creating searchable reference guides for complex development patterns
