skill-creator

To create and package modular AI skills, use a scaffolded workflow that generates a standardized SKILL.md frontmatter template with required name and description fields, ensuring consistent structure and discoverability for distribution.

The best way to structure an Agent Skill for progressive disclosure is to scaffold directories for scripts, references, and assets, guided by standardized frontmatter conventions that enforce consistent naming and descriptive descriptions.

You generate a SKILL.md template automatically by running an initialization script from your command line, specifying the skill name, output directory, and optional resource scaffolding flags to produce the required frontmatter instantly.

Yes, you can restructure an existing skill and update its documentation conventions by applying a standardized scaffold that refines skill quality across scripts, references, and assets while enforcing consistent naming.

The skill packaging workflow requires PyYAML as a dependency to process the frontmatter templates and manage the standardized metadata needed to produce ready-to-package .skill files.

You should not use a scaffolded template for skill creation if your project requires a highly custom directory structure that deviates from standardized frontmatter, or if you do not need to package the skill for modular distribution.