MCP server
Living Brain hosts a Model Context Protocol server, so AI agents can capture knowledge, search pages, and manage brains without any code on your side.
It uses the Streamable HTTP transport and calls the production API at
https://api.livingbrain.com.
Connect a client
claude mcp add --transport http livingbrain \ https://livingbrain.fernmcp.dev/livingbrain \ --header 'x-api-key: YOUR_API_KEY' \ --header 'x-subject-id: YOUR_SUBJECT_ID'
Run it in a terminal. Claude Code picks the server up on its next start.
Replace YOUR_API_KEY and YOUR_SUBJECT_ID with your own values (see Authenticate).
Any other MCP client that supports remote HTTP servers with custom headers works the same way.
Authenticate
The server forwards the same two headers as the SDKs and CLI. Configure your client to send both on every request:
See Authentication for how the key and subject relate. Treat the config file like any other place you store the key.
Tools
Each API operation is one tool, named <namespace>_<method> after the SDK methods and grouped
into toolsets by namespace:
Tool arguments follow the API Reference: path, query, and body parameters become tool inputs, and calls return the same JSON responses.