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

# Installation

> Connect the Xpoz MCP server to Claude Code, Claude Desktop, Gemini CLI, Cursor, N8N, and other clients

The Xpoz MCP server is hosted at `https://mcp.xpoz.ai/mcp`. You need an access key to authenticate -- get one from your [Xpoz dashboard](https://xpoz.ai).

<Tip>
  **Start instantly — no account needed.** Generate a free token and use it in place of `YOUR_API_KEY` below:

  ```bash theme={null}
  curl -s -X POST https://api.xpoz.ai/api/trial/token \
    -H "Content-Type: application/json" \
    -d '{"source": "docs"}' | jq -r .data.accessKey
  # -> TRIAL...  (valid 5 days, all read tools across 4 platforms)
  ```

  This returns a preview of up to 5 results per call. To get full data, pagination, and CSV export, [create a free account](https://xpoz.ai/get-token) — no credit card required.
</Tip>

<Tabs>
  <Tab title="Claude Code">
    Run a single command in your terminal:

    ```bash theme={null}
    claude mcp add xpoz-mcp https://mcp.xpoz.ai/mcp -t http -H "Authorization: Bearer YOUR_API_KEY"
    ```

    Claude Code will automatically discover all 48 tools on the next conversation.

    <Tip>
      You can verify the connection by asking Claude: "What Xpoz tools are available?"
    </Tip>
  </Tab>

  <Tab title="Claude Desktop">
    1. Open Claude Desktop
    2. Go to **Settings** > **Developer** > **Edit Config**
    3. Add the following to your configuration file:

    ```json theme={null}
    {
      "mcpServers": {
        "xpoz": {
          "url": "https://mcp.xpoz.ai/mcp",
          "headers": {
            "Authorization": "Bearer YOUR_API_KEY"
          }
        }
      }
    }
    ```

    4. Restart Claude Desktop

    <Note>
      Replace `YOUR_API_KEY` with your actual Xpoz access key. You can find it in your [Xpoz dashboard](https://xpoz.ai).
    </Note>
  </Tab>

  <Tab title="Gemini CLI">
    Add the MCP server to your Gemini CLI settings file (`~/.gemini/settings.json`):

    ```json theme={null}
    {
      "mcpServers": {
        "xpoz": {
          "uri": "https://mcp.xpoz.ai/mcp",
          "headers": {
            "Authorization": "Bearer YOUR_API_KEY"
          }
        }
      }
    }
    ```

    Restart Gemini CLI to pick up the new configuration.
  </Tab>

  <Tab title="Cursor">
    1. Open Cursor Settings (`Cmd+,` / `Ctrl+,`)
    2. Navigate to **MCP** section
    3. Click **Add new MCP server**
    4. Configure:
       * **Name**: `xpoz`
       * **Type**: `http`
       * **URL**: `https://mcp.xpoz.ai/mcp`

    Alternatively, add to your `.cursor/mcp.json`:

    ```json theme={null}
    {
      "mcpServers": {
        "xpoz": {
          "url": "https://mcp.xpoz.ai/mcp",
          "headers": {
            "Authorization": "Bearer YOUR_API_KEY"
          }
        }
      }
    }
    ```

    <Note>
      Set the `Authorization` header with your bearer token. Cursor supports HTTP transport natively.
    </Note>
  </Tab>

  <Tab title="N8N">
    1. Create an **AI Agent** node in your workflow
    2. Add a tool: **MCP Client Tool**
    3. Configure the MCP Client Tool:

    | Setting                        | Value                     |
    | ------------------------------ | ------------------------- |
    | **Endpoint**                   | `https://mcp.xpoz.ai/mcp` |
    | **Server Transport**           | HTTP Streamable           |
    | **Authentication Type**        | Bearer Auth               |
    | **Credential for Bearer Auth** | Your Xpoz access key      |

    4. Click **Refresh** to load the available tools
    5. Select which tools to include in your agent's toolset

    <Tip>
      If you can see the tools list after clicking Refresh, you are successfully connected to the MCP server.
    </Tip>
  </Tab>
</Tabs>

## Getting Your Access Key

1. Sign up or log in at [xpoz.ai](https://xpoz.ai)
2. Navigate to your dashboard
3. Copy your access key

You can also retrieve your access key programmatically using the `getUserAccessKey` tool if you are authenticated via OAuth, or check its status with `checkAccessKeyStatus`.

## Verify the Connection

After setup, test the connection by asking your AI agent to run a simple query:

```
Search for Twitter user @elonmusk
```

If the agent successfully calls `getTwitterUser` and returns profile data, your connection is working.

## Next Steps

<CardGroup cols={2}>
  <Card title="Browse Tools" icon="wrench" href="/mcp/tools/overview">
    See all 48 tools organized by platform.
  </Card>

  <Card title="Response Modes" icon="sliders" href="/mcp/response-modes">
    Choose between fast, paging, and CSV response modes.
  </Card>
</CardGroup>