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

# MCP Server

> Connect AI tools to the Lumx documentation using the Model Context Protocol

The Lumx documentation provides a [Model Context Protocol](https://modelcontextprotocol.io/) (MCP) server that lets AI tools search and retrieve content directly from these docs. This means you can ask questions about the Lumx API from tools like Claude, Cursor, or VS Code and get answers grounded in the official documentation.

## Server URL

```
https://docs.lumx.io/mcp
```

## Setup

<Tabs>
  <Tab title="Claude Code">
    Run the following command in your terminal:

    ```bash theme={null}
    claude mcp add --transport http lumx-docs https://docs.lumx.io/mcp
    ```
  </Tab>

  <Tab title="Cursor">
    Add the following to your `.cursor/mcp.json` file:

    ```json theme={null}
    {
      "mcpServers": {
        "lumx-docs": {
          "url": "https://docs.lumx.io/mcp"
        }
      }
    }
    ```
  </Tab>

  <Tab title="VS Code">
    Add the following to your `.vscode/mcp.json` file:

    ```json theme={null}
    {
      "servers": {
        "lumx-docs": {
          "type": "http",
          "url": "https://docs.lumx.io/mcp"
        }
      }
    }
    ```
  </Tab>

  <Tab title="Codex">
    Add the following to your `~/.codex/config.toml` file:

    ```toml theme={null}
    [mcp_servers.lumx-docs]
    url = "https://docs.lumx.io/mcp"
    ```

    Restart the CLI to pick up the new server. Codex Cloud doesn't currently expose MCP configuration — use the CLI for live docs lookup.
  </Tab>

  <Tab title="Claude Web">
    1. Go to **Settings** > **Connectors** in [Claude](https://claude.ai)
    2. Select **Add custom connector**
    3. Enter `Lumx Docs` as the name
    4. Paste `https://docs.lumx.io/mcp` as the URL
  </Tab>
</Tabs>

## Available tools

The server exposes two tools to connected AI clients. Your client picks the right one based on the question, so you don't need to call them directly.

* `search_lumx`: semantic search across the docs. Returns matching pages with titles, snippets, and direct links. Best for "where do I find X?" questions.
* `query_docs_filesystem_lumx`: shell-style access to the raw `.mdx` and OpenAPI files using `rg`, `head`, `cat`, and `tree`. Best for fetching full request schemas, listing endpoints, or grepping for specific patterns.