generated: '2026-09-11'
method: searched
source: https://www.socialfetch.dev/llms.txt
type: LLMsTxt
files:
  - path: llms/social-fetch-llms.txt
    url: https://www.socialfetch.dev/llms.txt
    status: 200
    description: Curated discovery + tool-selection index (11 KB).
  - path: llms/social-fetch-agents.txt
    url: https://www.socialfetch.dev/agents.txt
    status: 200
    description: >-
      AI agent / crawler access file — onboarding, auth rules, cursor-pagination
      rules, crawl allowlist, and per-platform endpoint counts (last updated
      2026-09-08).
  - url: https://www.socialfetch.dev/llms-endpoints.txt
    status: 200
    description: Full 206-operation catalog, each linking that route's markdown page (30 KB, not stored).
  - url: https://www.socialfetch.dev/llms.json
    status: 200
    description: >-
      Structured JSON inventory — parameters, pagination, outcomes, credits, SDK
      mapping (436 KB, not stored; llms-full not committed per repo policy).
notes: >-
  Social Fetch ships an unusually complete agent file set: llms.txt, agents.txt,
  a per-operation llms-endpoints.txt, per-platform llms-{platform}.txt files, and
  a structured llms.json. Docs pages also serve markdown twins (append .mdx, or
  Accept: text/markdown with Vary: Accept). The large llms.json / llms-full dumps
  are intentionally not committed (see .gitignore *-llms-full.txt).