skill-creator

To package AI skills, you utilize utility scripts like `package_skill.py` to validate the structure and bundle the `SKILL.md`, `scripts/`, and `references/` directories into a self-contained extension ready for deployment.

The anatomy of an AI skill consists of a modular, self-contained structure featuring a core `SKILL.md` file alongside organized `scripts/`, `references/`, and `assets/` directories to define capabilities and resources.

You initialize a new AI skill by running the provided `init_skill.py` script, which scaffolds the essential directory structure and files, allowing you to immediately begin defining your custom AI capabilities.

Building AI skills requires developer scripting knowledge to create functional utility scripts and understand the modular packaging process, though the provided tooling automates initial scaffolding and final validation.

The best way to structure an AI skill for custom document analysis is to define the logic in `SKILL.md`, place executable code in `scripts/`, and organize supporting reference documents within the `references/` directory.

Yes, you can iterate on existing AI skills by modifying the modular components within the established structure and re-running the packaging scripts to validate changes and update the distributed extension.