Connecting to Miro MCP

📘

This is a beta release of our MCP: sign up at https://miro.com/ai/mcp/.

It is intended for evaluation and feedback purposes. The feature set and user experience may change in future releases.

This guide walks you through connecting your AI tool to Miro using the Model Context Protocol (MCP).

Prerequisites

Before you begin, ensure that:

• You have a supported AI tool with MCP client capabilities.
• You have appropriate access permissions to your Miro workspace.
• Your tool is updated to support OAuth 2.1 authentication.

Configuration steps

1. Open your AI tool's settings.
2. Navigate to the MCP configuration section.
3. Add one of the following configurations based on your MCP client's capabilities:

For MCP clients with direct connection support:

{
  "mcpServers": {
    "miro-mcp": {
      "url": "https://mcp.miro.com/", 
      "disabled": false,
      "autoApprove": []
    }
  }
}

For MCP clients that require command-line connection handling:

{
  "mcpServers": {
    "miro": {
      "command": "npx",
      "args": [
        "mcp-remote",
        "https://mcp.miro.com"
      ],
      "disabled": false,
      "autoApprove": []
    }
  }
}

The first configuration is for clients that can connect directly to the MCP server URL. The second configuration is for clients that need to use a command-line tool as an intermediary for handling the connection, which might be necessary in certain environments or for clients without built-in remote connection capabilities.

📘

A note on using mcp-remote or similar open-source tools

Please be aware that mcp-remote is a third-party, open-source tool available on NPM. While it enables command-line connection handling, we do not maintain this package and cannot be held responsible for its security. We advise using it at your own discretion and understanding the potential risks of running third-party code.

Security and rate Limits

• Authentication uses OAuth 2.1 with dynamic client registration.
• Standard API rate limits apply to all operations.
• Enterprise security compliance is maintained.
• Access is governed by user permissions.

Troubleshooting issues

If you're experiencing connection issues, try these solutions:

1. Check MCP client support: Verify your AI tool supports MCP clients and can connect to MCP servers. Supported tools include Cursor, Lovable, and GitHub Copilot.

2. Verify remote server support: Some AI tools have MCP clients but don't support remote connections. For these cases:

   • Check your tool's documentation for remote connection support.
   • Ensure you're using the latest version of your AI tool.
   • Verify your network can reach the MCP server.

3. Use MCP Inspector: For detailed connection diagnostics, use MCP Inspector.

4. Request MCP support: If your AI tool doesn't support MCP, contact the tool's developers to request MCP server connection support.