Agent integration

.well-known/rig.json

A small site-level descriptor an agent can fetch to learn what rig is and where things live.

The Hub serves a small JSON file at /.well-known/rig.json. It's intended for agents that want to discover rig from any web context.

What's in it

A small, stable descriptor with pointers to the CLI install, the docs, the Hub API base, and a brief description of what rig is.

The site also advertises this descriptor via HTTP Link headers on the root response, plus a service-doc link pointing at /docs.

Why it's useful

When an agent lands on userig.xyz (or follows a link to a rig page), it can fetch the descriptor and have a structured starting point — no scraping required.

Browser agents reading the Hub can use this to:

  • Find the CLI install command without searching the docs.
  • Identify the docs base URL programmatically.
  • Discover the API base URL for direct queries.

Example use

curl -s https://userig.xyz/.well-known/rig.json | jq

Or, in an agent:

1. Fetch /.well-known/rig.json.
2. Read the docs base URL.
3. Direct further questions to the docs.
4. Use the API base URL for programmatic queries (e.g. "what rigs match X?").

Caveats

  • The descriptor is informational. It does not authenticate; it does not authorise.
  • The schema is stable but not yet versioned strongly. Treat unknown keys as additive (forward-compatible).

See also