Integrations

How to connect Tunio MCP

Step-by-step guide to connecting Tunio MCP in Codex/Cursor and Claude Desktop.

Tunio MCP gives your AI assistant direct access to your Tunio streams and actions, so it can read current tracks, list stations, and run MCP tools.

Requirements

Never expose your API key in screenshots, repositories, or chats. If a key is leaked, revoke it and create a new one.

Connect in Codex / Cursor

  1. Open MCP server settings

    Go to the MCP server add/edit section in your AI client.

  2. Set command and arguments

    Use npx as the command and add arguments one by one as separate entries:

    BASH
    -y
mcp-remote
https://agent.tunio.ai/mcp
--header
Authorization: Bearer YOUR_API_KEY

  3. Save and refresh tools

    Save the MCP configuration and restart the client (or refresh tools) so Tunio tools appear.

Connect in Claude Desktop

The setup is the same as in Codex/Cursor.

  1. Add a new MCP Server

    Open MCP settings in Claude Desktop and create a new server.

  2. Fill command and args

    Use the same values:

    BASH
    command: npx
args:
  - -y
  - mcp-remote
  - https://agent.tunio.ai/mcp
  - --header
  - Authorization: Bearer YOUR_API_KEY

  3. Restart Claude Desktop

    After saving, restart the app and verify that Tunio tools are listed.

Verify the connection

Run a simple tool call such as list_stations or get_user_context.

If tools do not appear:

  • Check that npx is available in your environment.
  • Verify the API key.
  • Make sure args are added as separate items, not a single string.
  • Restart the AI client.

If a test call returns data (for example, your stream list), Tunio MCP is connected correctly.