> ## Documentation Index
> Fetch the complete documentation index at: https://docs.multivon.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# multivon-mcp

> MCP server that gives AI coding agents direct access to the multivon-eval and pdfhell evaluation surface.

`multivon-mcp` is a [Model Context Protocol](https://modelcontextprotocol.io) server that exposes 19 evaluation tools to any MCP-compatible agent — Claude Desktop, Claude Code, Cursor, Cline, OpenCode. The agent calls them by name; no copy-paste, no `python -c "..."`, no asking the agent to figure out the SDK calls.

When the agent is helping you build an LLM product, it can:

* Score a RAG output for hallucination without you writing the scaffolding
* Generate an adversarial PDF on demand to test your document AI
* Run the full pdfhell mini-suite against a model and analyse the results
* Produce a hash-chained audit pack for procurement diligence
* Discover the full evaluation capability catalog as JSON

## Why an MCP server (not just a CLI)

The MCP layer matters because the agent is the user. A CLI is the right shape for humans driving terminals; an MCP server is the right shape for an LLM that needs to *call* the tool mid-edit, get JSON back, and condition its next action on the result.

Three concrete wins:

1. **Tool discovery is free.** `eval_discover` returns the full 44-evaluator catalog as JSON. The agent never has to parse markdown to know what's available.
2. **JSON results, not pretty-printed.** Every tool returns `{"score": 0.0-1.0, "passed": bool, "reason": str, "threshold": float}`. The agent can branch on the score programmatically.
3. **Calibration is automatic.** Each evaluator uses the calibrated threshold for the configured judge — no asking the agent to guess at the right cutoff.

## Install

```bash theme={null}
pip install multivon-mcp
```

The bare install pulls `multivon-eval`, `pdfhell`, and the MCP SDK. Provider SDKs (`anthropic`, `openai`, `google-genai`) come along — bring your own API key in env.

<Tip>
  For one-off use without installing, `uvx multivon-mcp` works zero-setup.
</Tip>

## Next steps

* [Configuration](/mcp/configuration) — wire it into Claude Desktop, Claude Code, Cursor, or Cline.
* [Tool reference](/mcp/tool-reference) — every tool, every argument.
* [Agent recipes](/mcp/agent-recipes) — short patterns for common eval flows.