PageLeft 🍄

PageLeft searches copyleft and public-domain writing and code:
CC BY-SA, GFDL, AGPL/GPL, CC0, and Public Domain. Find relevant
passages with their source URL and detected license attached.

12384 pages indexed, 238172 text chunks. Copyleft and public-domain sources
are admitted; other licenses, including CC BY-ND and proprietary
licenses, are excluded by design.

What it does
  Given a natural-language query, it returns the chunks most relevant
  to the query, with source URL and detected license attached. A human
  or a coding agent can then inspect the source and its terms before
  quoting or adapting it. Each source retains its own license.

API
  GET  /api/search?q=<query>     search by natural language
  GET  /api/stats                index size
  GET  /api/leaderboard          contributor rankings
  POST /api/contribute/page      submit a URL for indexing

Example
  curl https://pageleft.cc/api/search?q=copyleft+ratchet

Use in a coding agent
  Any agent that can make HTTP requests can use PageLeft as a parts
  bin for free software and prose. Point the agent at /api/search,
  have it preserve the returned URL and license with any quoted or
  adapted text. Check the source terms when combining works; results
  are not all under one license.

  For Claude Code specifically:
    claude plugin marketplace add kimjune01/pageleft
    claude plugin install pageleft@pageleft

Source
  https://codeberg.org/kimjune01/pageleft   (Forgejo, no nonfree JS required)
  https://github.com/kimjune01/pageleft     (mirror)

Read more: https://www.june.kim/pageleft
Contribute: https://pageleft.cc/contribute
License: AGPL-3.0-or-later (PageLeft code); indexed sources retain their own licenses.