For AI clients
This documentation is built to be consumed by AI agents and tools, not just people. There are two standard ways to give a model access to it.
1. llms.txt (fetch-based)
We publish the llmstxt.org standard files. Point your model or agent at these:
- llms.txt — a curated index: every doc with a one-line summary and a link. Best when the model should pick what to read.
- llms-full.txt — the entire documentation corpus inline as one file. Best for dropping the whole context into a prompt or a RAG pipeline.
Both are plain text, regenerated on every build, and need no authentication.
2. MCP (live, tool-based)
Eve runs a Model Context Protocol server, so a connected client can read these docs as live resources and call Eve's tools — no copy- paste, always current.
- Connect an AI client — setup for Claude Code, Claude
Desktop, and other MCP clients (Workspace MCP at
/mcp, driver Connect MCP at/connect/mcp). - MCP server overview — authentication, tool discovery, and the documentation resources Eve exposes.
Which should I use?
| You want to… | Use |
|---|---|
| Drop docs into a prompt or RAG index | llms-full.txt |
| Let a model browse and fetch pages on demand | llms.txt |
| Give an agent live, authenticated access to Eve + current docs | MCP |