DAAF Skill Authoring Framework
Standardizes the creation, auditing, and optimization of agent skills within the DAAF ecosystem.
The Skill Authoring framework provides a comprehensive set of standards and best practices for developing high-quality DAAF (Data-Agile Agent Framework) skills. It guides developers through the entire lifecycle of skill creation—from defining YAML frontmatter and naming conventions to implementing advanced progressive disclosure patterns that optimize token consumption. By using this skill, authors can ensure their specialized agent capabilities are correctly discovered, triggered accurately by Claude, and structured for maximum performance using standardized directory layouts for scripts, references, and assets.
주요 기능
01Standardized directory organization for bundled resources
021,851 GitHub stars
03Comprehensive auditing checklists for debugging and optimization
04YAML frontmatter validation and metadata standards
05Decision tree patterns for structured agent navigation
06Three-level progressive disclosure for token budget management
사용 사례
01Designing new domain-specific capabilities for AI agents
02Auditing existing skills to resolve under-triggering or over-triggering issues
03Optimizing large documentation sets into performance-ready AI instructions