mcp

The TagoIO MCP Server lets AI models interact directly with your TagoIO account, providing contextual access to devices, data, and platform resources. It supports features like device information and real-time data access, data analysis operations (sums, averages, and reports), platform integration (retrieving users, actions, analysis scripts, and account statistics), and AI-assisted generation of TagoIO Analysis scripts with proper context. It also offers development support such as debug assistance and tag relationship analysis to streamline experimentation and integration workflows.

To use the server, deploy it with your TagoIO token and target API endpoint. The server exposes an API surface that an AI agent can query to fetch device data, run analyses, or generate code snippets. Provide the appropriate token with the minimal necessary permissions for the resources you intend to access (External analysis tokens are recommended for security).

Once running, you can instruct the AI to perform tasks like fetching device metadata, querying recent data, generating an Analysis script for a given dataset, or retrieving account statistics. The environment variable TAGOIO_API lets you switch between US and EU regions, ensuring requests are sent to the correct regional endpoint.

Prerequisites:

• Node.js 18+ installed on your machine
• A TagoIO account with a valid profile or analysis token
• Internet access to pull the MCP server package

Step 1: Create the MCP configuration file

• Create or update a configuration file (for example mcp_config.json) with the following content:

{
  "mcpServers": {
    "@tago-io/mcp": {
      "command": "npx",
      "args": ["-y", "@tago-io/mcp-server"],
      "env": {
        "TAGOIO_TOKEN": "YOUR-TOKEN",
        "TAGOIO_API": "https://api.us-e1.tago.io"
      }
    }
  }
}

Step 2: Install prerequisites (if needed) and run

• Ensure Node.js is installed and in your PATH.
• Run the MCP server using the configuration above. The recommended approach is to use the provided npx command, which will fetch and execute the MCP server package with the specified environment variables.

# Example: run the server using the configuration embedded in the mcp_config.json (adjust as needed for your setup)
npx -y @tago-io/mcp-server

Step 3: Verify and use

• After starting, the MCP server will be reachable according to the package defaults. Use your AI platform or IDE to connect to the server and begin issuing requests such as retrieving devices, data, or generating Analysis scripts.
• If you operate in the EU region, update TAGOIO_API to https://api.eu-w1.tago.io.

Step 4: Security considerations

• Use an Analysis token with restricted permissions where possible.
• Do not commit tokens or secrets to version control.
• Consider rotating tokens periodically and limiting access to the minimum required scopes.

Tips and common issues:

• Environment variables: TAGOIO_TOKEN should be a token with the necessary permissions for the intended resources. TAGOIO_API should reflect your region (US or EU). If you switch regions, ensure both the API URL and any region-specific endpoints are updated.
• Token scope: Prefer an Analysis token with limited permissions to reduce risk; if using a Profile token, understand it grants broader access.
• If you encounter authentication errors, verify token validity and that the token has not expired or been revoked.
• When using external AI tools, ensure they can reach the MCP server endpoints and that any required prompts or prompts templates are correctly formatted for generating or requesting code snippets.
• If you need to run the server inside a development IDE or via a one-click installer, you can supply the same configuration JSON in your IDE's MCP integration settings.
• EU vs US endpoints: The server supports both; ensure you point to the correct API depending on your TagoIO account region.