skill-creator

To create AI skills, follow a structured process covering skill anatomy, conciseness, and progressive disclosure. You build modular units containing specialized knowledge, workflows, and tool integrations by planning, initializing, and iterating on the design.

Progressive disclosure in agent skill development is a core design principle that organizes information hierarchically. It ensures AI agents receive necessary specialized knowledge contextually without being overwhelmed, keeping the skill concise and self-contained.

The best way to organize agent skills involves using a specific anatomy: a SKILL.md file for core instructions, alongside scripts/, references/, and assets/ directories. This structure ensures modular, self-contained units with clear workflows.

Yes, you can use structured documentation to integrate external APIs into a skill. By following the creation process, you organize necessary scripts and references within the skill structure to interact seamlessly with specific company APIs.

The core principles for writing a concise SKILL.md include focusing on freedom of expression and progressive disclosure. These principles guide how to effectively extend AI capabilities while keeping the documentation efficient and modular.

The step-by-step process for updating AI skills involves planning, initializing, editing, packaging, and iterating on the design. This ensures existing skills remain well-structured, effective, and efficient in extending AI capabilities.