> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.rebateright.com.au/mcp/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.rebateright.com.au/_mcp/server.

# MCP server

These docs host a Model Context Protocol (MCP) server. Connect it once, and your AI coding assistant can search every endpoint, schema, and guide in this documentation while it writes your integration.

```bash
claude mcp add --transport http rebateright-docs https://docs.rebateright.com.au/_mcp/server
```

## Why connect it

An assistant working from training data alone guesses at endpoint paths, request fields, and error codes. The MCP server replaces guessing with lookup: it gives your assistant a `searchDocs` tool it calls whenever it needs a fact from this documentation, and every answer comes from the docs as published today, with source URLs back to the page it quoted.

| Field              | Value                                                               |
| ------------------ | ------------------------------------------------------------------- |
| **Endpoint**       | `https://docs.rebateright.com.au/_mcp/server`                       |
| **Transport**      | Streamable HTTP                                                     |
| **Authentication** | None. The server is public and read-only.                           |
| **Tools**          | `searchDocs(query)`: returns relevant doc passages with source URLs |

The server is generated from the same source as these docs, so it updates with every release.

## Connect your client

#### Claude Code

Register the server from a terminal:

```bash
claude mcp add --transport http rebateright-docs https://docs.rebateright.com.au/_mcp/server
```

Run `/mcp` inside Claude Code to confirm `rebateright-docs` shows as connected.

#### Cursor

Add the server to `.cursor/mcp.json` in your project, or `~/.cursor/mcp.json` for all projects:

```json
{
  "mcpServers": {
    "rebateright-docs": {
      "url": "https://docs.rebateright.com.au/_mcp/server"
    }
  }
}
```

#### VS Code

Add the server to `.vscode/mcp.json` in your workspace:

```json
{
  "servers": {
    "rebateright-docs": {
      "type": "http",
      "url": "https://docs.rebateright.com.au/_mcp/server"
    }
  }
}
```

#### Other clients

Any client that speaks MCP over streamable HTTP can connect. Register a remote server with:

| Field     | Value                                         |
| --------- | --------------------------------------------- |
| Name      | `rebateright-docs`                            |
| URL       | `https://docs.rebateright.com.au/_mcp/server` |
| Transport | Streamable HTTP                               |

No token, header, or environment variable is required.

## Try it

Ask your assistant a question it cannot answer from memory:

> Which RebateRight endpoint verifies a patient's Medicare card, and what does the request body look like?

A connected assistant calls `searchDocs`, reads the matching pages, and answers with the current endpoint path and schema, citing the docs it used. If it answers without citing these docs, check the connection.

## Other machine-readable formats

The MCP server suits agents that decide at runtime what to look up. For pipelines that want the docs as files instead, every machine-readable format these docs publish, OpenAPI specification included, is listed under [AI-Ready Documentation](/#ai-ready-documentation) on the home page.

> Connect Claude Code, Cursor, or any MCP client to a live search tool over these docs