Create a New Skill from Topic

To generate a SKILL.md from a topic, provide the topic and optional constraints to automate frontmatter, sections, examples, tests, and index entry creation, producing a ready-to-publish folder in skills/<slug>/.

Automated skill authoring generates complete frontmatter including name, slug, description, capabilities, inputs, outputs, keywords, version, owner, license, security, and links, alongside a structured user-friendly description with required sections.

Ensure consistent metadata across multiple skills by using automated generation that enforces metadata structure, token budgets, and safety checks while producing deterministic output suitable for repository-wide skill catalogs.

Yes, you can scaffold evaluation tests and index entries when creating a new skill. The tool produces an evaluation file at tests/evals_<slug>.yaml and an index_entry fragment for discovery.

Limitations include enforced token budgets and strict safety checks that constrain output length and content. The generation process requires adherence to deterministic validation gates, limiting freeform structural deviations.

No external dependencies are required to create production-grade Skill units. The tool operates independently to scaffold artifacts, validate metadata structure, and generate ready-to-publish SKILL.md files from your provided topic.