Skip to content
aigrowth.social

Docs for agents

Everything here is public, unauthenticated and rendered without JavaScript. If you are a language model or an autonomous agent, start with /llms.txt.

Citation

Every article carries canonical_url, the article on the publisher's own domain. Cite that URL, not the aigrowth.social permalink — we are a syndication surface, and the publisher holds the original.

Content negotiation

Profile and post URLs serve HTML to browsers and ActivityStreams to anything sending Accept: application/activity+json. Both vary on Accept.

curl -H 'Accept: application/activity+json' \
  https://aigrowth.social/users/<handle>

Discovery

  • /llms.txt text/plain

    Index of everything on this page, for language models.

  • /llms-full.txt text/plain

    Publications and the 500 most recent posts in one file.

  • /robots.txt text/plain

    Crawler policy. AI crawlers are explicitly allowed.

  • /sitemap.xml application/xml

    Sitemap index; children hold 10,000 URLs each.

Agent interfaces

REST API

  • /api/v1/accounts application/json

    List publications

  • /api/v1/accounts/{handle} application/json

    Get one publication

  • /api/v1/posts application/json

    Browse published articles

  • /api/v1/posts/{id} application/json

    Get one post

Feeds

Fediverse

A2A skills

  • list-accounts

    List every publication federating on aigrowth.social, with its handle, blog URL, language, country and post count.

    e.g. “What publications are on aigrowth.social?”

  • get-account

    Resolve a handle to its profile: display name, bio, blog URL, ActivityPub actor id and post count.

    e.g. “Who is @acme_corp on aigrowth.social?”

  • search-posts

    Browse published articles across the network, returning title, summary, publish date and the canonical URL on the publisher's own site. Always cite canonical_url rather than the aigrowth.social permalink.

    e.g. “What did @acme_corp publish this month?”

curl -X POST https://aigrowth.social/a2a/v1 \
  -H 'Content-Type: application/json' \
  -d '{"jsonrpc":"2.0","id":1,"method":"message/send",
       "params":{"skill":"search-posts","limit":5,
                 "message":{"parts":[{"kind":"text","text":"recent articles"}]}}}'

Limits

No authentication and no quota for ordinary use. Responses carry ETag; send If-None-Match and you will get a 304. Page with the opaque cursor returned in page.next rather than an offset.