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

# Connect your AI assistant

> Add the Porto documentation to your AI tools through the hosted MCP server.

<Callout icon="vault" color="#ac96ff">
  You're viewing **Porto** documentation — a distinct product from Anchorage Digital's core custody platform. [Go to Anchorage Digital docs →](/knowledge-base/build-with-ai/overview)
</Callout>

Connect our documentation to your AI assistant so it can answer questions and generate integration code grounded in the current docs. Anchorage Digital hosts a Model Context Protocol (MCP) server for this site.

## What the docs MCP server does

The MCP server lets your assistant search these docs and read pages on demand. It's read-only documentation knowledge—it can't access your Porto wallet, keys, or funds, and it can't move assets. To take actions, your integration calls the REST API with your own scoped keys.

| Server             | URL                                     | Use for                                                       |
| :----------------- | :-------------------------------------- | :------------------------------------------------------------ |
| Public docs        | `https://docs.anchorage.com/mcp`        | Public documentation                                          |
| Authenticated docs | `https://docs.anchorage.com/authed/mcp` | Documentation restricted to your account, after OAuth sign-in |

## Quick install

The fastest path is the one-line installer, which detects your installed tools:

```bash theme={null}
npx add-mcp https://docs.anchorage.com/mcp
```

You can also open the AI menu at the top of any docs page and choose an install option, such as Cursor, VS Code, or Copy MCP.

## Set up by tool

### Claude Code

```bash theme={null}
claude mcp add --transport http anchorage-docs https://docs.anchorage.com/mcp
```

### Claude Desktop and Claude web

<Steps>
  <Step title="Open connector settings">
    In Settings, select **Connectors**, then select **Add custom connector**.
  </Step>

  <Step title="Add the server">
    Enter a name and the URL `https://docs.anchorage.com/mcp`, then save.
  </Step>
</Steps>

### Cursor

Add the server to `mcp.json`:

```json theme={null}
{
  "mcpServers": {
    "anchorage-docs": {
      "url": "https://docs.anchorage.com/mcp"
    }
  }
}
```

### VS Code (GitHub Copilot)

Add the server to `.vscode/mcp.json`:

```json theme={null}
{
  "servers": {
    "anchorage-docs": {
      "type": "http",
      "url": "https://docs.anchorage.com/mcp"
    }
  }
}
```

### ChatGPT and other clients

In any client that supports custom MCP connectors, add a new connector with the URL `https://docs.anchorage.com/mcp`. To include documentation restricted to your account, use `https://docs.anchorage.com/authed/mcp` and sign in when prompted.

<Note>
  The docs MCP server provides documentation only. It never has access to your API keys or the ability to move assets. See [AI security and guardrails](/knowledge-base/porto/build-with-ai/security).
</Note>

## Next steps

With your assistant connected, start from a [ready-made prompt](/knowledge-base/porto/build-with-ai/prompts), or review the [security guardrails](/knowledge-base/porto/build-with-ai/security) before you write code that moves assets.


## Related topics

- [Connect your AI assistant](/knowledge-base/build-with-ai/connect-ai-assistant.md)
- [Build with AI](/knowledge-base/build-with-ai/overview.md)
- [Use these docs with AI](/knowledge-base/platform/users/using-ai.md)
- [Sandbox testing guide](/knowledge-base/platform/developers/sandbox/sandbox-testing-guide.md)