skill-creator

To create reusable AI skills, use a guided workflow for discovery, initialization, and validation that generates compliant YAML frontmatter, hyphen-case naming, and proper directory scaffolding. This ensures your skill package is structured correctly for reliable reuse by future AI agents.

Progressive disclosure in skill authoring is a structural pattern that keeps SKILL.md concise by moving complex logic into bundled scripts and references. Use it when designing specialized workflows or domain knowledge integrations to maintain operational efficiency without overwhelming the main instruction file.

You validate YAML frontmatter and naming rules by running a built-in validator tool that checks your skill package for compliance. This validation process ensures hyphen-case naming conventions and required metadata fields are correctly formatted before delivery.

Skill creation requires the YAML dependency for frontmatter validation and metadata compliance. The workflow handles directory scaffolding, script generation, and reference organization natively, ensuring AI agent tooling packages remain self-contained without needing additional external libraries.

The best way to structure a new skill is to follow a guided workflow that separates core instructions in SKILL.md from operational scripts and reference templates. This packaging approach ensures specialized workflows like document automation remain concise and reliable for future AI agents.

You should avoid bundling scripts and references when a skill's core logic can be fully expressed concisely within SKILL.md itself. Over-bundling unnecessary components violates progressive disclosure patterns and complicates validation, reducing the skill's reliability for AI agent reuse.