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

# Connect via MCP

> Use CutPro inside Claude, ChatGPT, and other AI tools via the Model Context Protocol

**MCP (Model Context Protocol)** is an open standard that connects AI tools to external services. CutPro offers two ways to connect, with different purposes:

<CardGroup cols={2}>
  <Card icon="book-open" title="Documentation MCP">
    Lets the AI **answer questions** about CutPro by reading the documentation. Read only.
  </Card>

  <Card icon="wand-sparkles" title="API MCP (actions)">
    Lets the AI **run the flow**: analyze, clip, render, and download, using your API key.
  </Card>
</CardGroup>

## Documentation MCP

A search server hosted on the docs themselves, at `https://cut.pro/docs/mcp`. With it, the AI searches and reads the pages to answer questions about the product. It does not run actions.

<Tabs>
  <Tab title="Claude">
    In **Settings → Connectors → Add custom connector**, enter the URL `https://cut.pro/docs/mcp`.
  </Tab>

  <Tab title="Cursor / VS Code">
    Add the MCP server with the URL `https://cut.pro/docs/mcp` in your editor's MCP settings.
  </Tab>

  <Tab title="ChatGPT">
    In **Settings → Connectors**, add a connector pointing to `https://cut.pro/docs/mcp`.
  </Tab>
</Tabs>

## API MCP (actions)

This server exposes **all v1 API endpoints** as tools (34 in total), covering workspace, balance, videos and uploads, clipping, clips, templates, renders, publishing, and connections. This way the AI runs the flow end to end, from `analyze_video` to `create_post`.

<Note>
  API access requires the **Pro** plan. Generate a key at [API keys](https://cut.pro/studio/me/api-keys) and use it as `CUTPRO_API_KEY`.
</Note>

### Claude Code

```bash theme={null}
claude mcp add cutpro \
  --env CUTPRO_API_KEY=YOUR_KEY \
  -- npx -y @cutpro/mcp
```

### Claude Desktop, Cursor, Windsurf, VS Code

They all use the same MCP server configuration format:

```json theme={null}
{
  "mcpServers": {
    "cutpro": {
      "command": "npx",
      "args": ["-y", "@cutpro/mcp"],
      "env": { "CUTPRO_API_KEY": "YOUR_KEY" }
    }
  }
}
```

<Tip>
  Multi-workspace key? Also add `CUTPRO_WORKSPACE_ID` to the `env`.
</Tip>

### Usage example

Once connected, just ask in natural language:

> "Analyze this YouTube video, generate the clips, and send me the 3 highest-rated ones already rendered."

The AI chains the tools on its own: `analyze_video` → `submit_clipping` → `get_submission` (until done) → `list_clips` → `render_clip` → `get_render` → `get_render_download`.

### Remote and web (ChatGPT, Claude.ai)

The same server runs in **remote HTTP** mode (self-host) with full **OAuth 2.1** (metadata discovery, dynamic client registration, PKCE, and tokens). To connect:

<Steps>
  <Step title="Add the connector">
    In ChatGPT (Connectors) or Claude.ai (Settings → Connectors → Add custom connector), enter the server URL: `https://mcp.cut.pro`.
  </Step>

  <Step title="Authorize with your key">
    The client opens a consent page. Paste your [API key](https://cut.pro/studio/me/api-keys). The key is validated and access is authorized.
  </Step>

  <Step title="Done">
    The client can now use CutPro's tools in conversations.
  </Step>
</Steps>

<Card icon="terminal" href="/docs/en/api-reference/quickstart" horizontal>
  Prefer to call the API directly? See the **Quickstart** with `curl` examples.
</Card>