gteam-skill-builder

To scaffold a new specialist from GitHub or OpenAPI specifications, the skill automates folder structure creation, SKILL.md.tmpl generation, and reference harvesting to output structured markdown into the specialists directory.

Building reference libraries from documentation sites is automated by crawling static docs via WebFetch, checking llms.txt, and synthesizing content into clean, structured markdown reference files for specialists.

Scraping API documentation without hitting rate limits is managed by preferring the Skill-Seekers API first, then falling back to WebFetch or curl with a 3s rate limit to throttle requests during reference harvesting.

Yes, this scaffolding tool works with Swagger specs and llms.txt by checking for llms.txt during source discovery and parsing OpenAPI or Swagger specifications to generate structured markdown references.

WebFetch may fail on certain documentation sites during reference harvesting when static content is unavailable, requiring a fallback to curl with a 3s rate limit to retrieve the reference data.