lookerctl

You can use this MCP server to expose an AI-enabled interface for Looker workflow automation. It enables clients to interact with Looker-related tools through a Model Context Protocol (MCP) server, supporting local development, AI-assisted analysis, and automated workflows in a scalable way.

How to use

Start the MCP server locally to enable AI agents to talk to Looker tooling. The server is invoked via your CLI with a simple command and offers a set of endpoints for managing Looker workflows, running analyses, and coordinating AI-driven tasks.

How to install

Prerequisites: you need a working Node.js and npm/yarn environment to install and run MCP utilities.

1. Install dependencies for the MCP server tooling.

2. Ensure you have access credentials configured for your Looker instance if you plan to connect to Looker APIs.

3. Install or prepare the MCP server package as instructed by your project setup and ensure the executable entry point is available on your PATH.

4. Run the MCP server using the official start command shown in the examples.

MCP server configuration and usage

The MCP server is started as a local process and can be integrated into clients like Claude Desktop or other AI agents. A typical setup shows starting the server and registering it in an AI client configuration.

{
  "mcpServers": {
    "looker": {
      "type": "stdio",
      "command": "lookerctl",
      "args": ["mcp"],
      "env": {
        "LOOKERSDK_BASE_URL": "https://your-instance.looker.com",
        "LOOKERSDK_CLIENT_ID": "your_client_id",
        "LOOKERSDK_CLIENT_SECRET": "your_client_secret"
      }
    }
  }
}

Examples and workflows

Use the MCP server to enable AI-driven workflows such as exporting LookML, validating configurations locally, and analyzing usage patterns. The server supports programmatic control via the MCP interface and can be paired with AI agents to perform tasks like dependency analysis and optimization.

Security and access

Protect access to the MCP server by limiting network exposure and using proper authentication for Looker API interactions. Treat credentials as secrets and rotate them regularly according to your security practices.

Troubleshooting

If the MCP server fails to start, verify that the command and arguments are correct, ensure the process has appropriate permissions, and confirm environment variables for Looker access are set correctly. Check that the client configuration references the MCP server with the proper name and path.

Notes

This MCP server integration is designed to work with AI agents that need programmatic access to Looker tooling. It exposes a manageable interface for command execution, data analysis, and workflow orchestration in development and production environments.