Looker Extension Onboarding

Step 1: Configure Connection

To use the Looker extension, you need to configure your Looker instance URL and credentials.

You can return to this onboarding at any time by running the Looker: Show Onboarding Walkthrough command from the Command Palette.

* Indicates required fields

The URL of your Looker instance.
Your API3 client ID
Your API3 client secret
Your Looker OAuth Client ID (client_guid)

Step 2: Select Project

Select the Looker project associated with this repository.

Step 3: Run MCP Server

Looker provides a built-in Managed MCP server (/mcp) that requires no local installation or command-line server process.

If your Looker instance has Managed MCP enabled, click Next to verify connection. If disabled, ask your Looker admin to enable MCP in Looker Admin settings (documentation).

Run the local MCP server in your terminal by copying the command below. Leave it running in the background.

Specify a remote URL to connect to an external server instead of running it locally.

Step 4: Looker Skills

The Looker extension auto-installs Looker skills within the project to help agents write LookML correctly.

You should select .claude/skills if you will be working with Claude agents.

These skills are managed by the extension, and you should not modify them manually as they may be overwritten by updates.

Step 5: Configure IDE MCP Client

Connect your Gemini CLI agent to the extension's MCP proxy. This will allow the agent to access Looker tools.

Click the button below to automatically update your config file.