write-a-skill

Scaffolding a new skill requires a valid SKILL.md frontmatter with name and description fields, plus populated corresponding directories, to generate both canonical and wrapper skill structures following project conventions.

Canonical and wrapper skill scaffolding is an automation process that generates a core implementation directory and a linked Claude wrapper directory, ensuring teams follow consistent project conventions when adding new AI capabilities.

Skill scaffolding prompts you to update the readme or index to include the newly created skill, ensuring the project documentation reflects the added canonical implementation and Claude wrapper.

To create a new skill using this scaffold template, you need a valid SKILL.md frontmatter containing the name and description, and the corresponding directories must be populated before the automation can generate the wrapper structure.

You should use a skill scaffold template instead of manual setup when your team needs a consistent, repeatable starter kit for adding new capabilities, ensuring both canonical implementations and Claude wrappers are prepared and linked correctly.