The "feed Cursor your library docs" workflow

Step one: find the documentation URL for whatever library Cursor is hallucinating against (https://library.dev/docs/getting-started, /api/v2, etc.). Step two: convert the URL — and ideally each major sub-page — to Markdown. Step three: save as docs/library-name.md at the root of your repo. Cursor's indexer immediately treats Markdown files as searchable, retrievable context.

Step four: add a one-line .cursorrules entry: "When using library X, consult docs/library-name.md for accurate API signatures." Cursor will then autonomously read the relevant section of the docs before writing code that uses the library.

Why URLs (not PDFs, not pasted text)

Library documentation lives on the web — versioned, searchable, regularly updated. Pasting docs into Cursor chat works for one task, then evaporates. Converting URLs to workspace Markdown gives you a snapshot you control: re-convert when the docs update, version-control the result, share with your team via the repo.