skill-creator

To create an AI skill, follow a structured creation process covering skill anatomy, core principles, and component organization. This involves defining the SKILL.md file, scripts, references, and assets to integrate specialized knowledge and external tools effectively.

Progressive disclosure in AI skill creation is a core principle that involves structuring skill components like references and scripts to provide information incrementally. This approach keeps the main SKILL.md concise while allowing the AI to access deeper specialized knowledge when needed.

The best way to structure AI skills is by organizing components into SKILL.md, scripts/, references/, and assets/ directories. This structure separates core instructions from executable scripts and reference materials, enabling efficient workflow design and tool integration.

You do not need advanced programming to build a skill with specialized knowledge, but structuring components like scripts and references requires developer-level familiarity. The skill creation process focuses on organizing knowledge and setting degrees of freedom for AI capabilities.

You can refine an existing AI workflow by iterating on its design and adjusting component organization. The skill creation guide supports improving existing skills by applying core principles like conciseness and progressive disclosure to enhance their capabilities.

You should avoid using a structured skill for AI development when the task requires simple, single-turn interactions without specialized knowledge or tool integrations. Skills are designed for complex workflows requiring progressive disclosure and structured component organization.