Guides the creation of high-performance Claude Code skills using Test-Driven Development (TDD) principles.
This skill provides a rigorous framework for authoring, testing, and optimizing specialized Claude Code skills to enhance agent behavior. By applying the Red-Green-Refactor cycle to documentation, it ensures that new skills are discoverable, token-efficient, and reliably followed. It places heavy emphasis on Claude Search Optimization (CSO), teaching authors how to write effective triggering conditions that prevent Claude from bypassing complex workflows or relying on hallucinated summaries.
Key Features
01TDD-based skill creation workflow (Red-Green-Refactor for documentation)
02Token efficiency guidelines for high-performance context management
030 GitHub stars
04Pressure-testing methodology using subagent baseline scenarios
05Claude Search Optimization (CSO) techniques for reliable skill discovery
06Standardized SKILL.md structure with YAML frontmatter templates
Use Cases
01Standardizing documentation across large-scale multi-agent repositories
02Encoding complex domain-specific architectural patterns for AI consistency
03Optimizing existing documentation to ensure Claude follows multi-step workflows
