Lightdash MCP Server

Give Claude Desktop read-only access to your Lightdash projects, charts, dashboards, and queries.

Setup

Step 1: Get a Lightdash Personal Access Token

1. Open your Lightdash instance in a browser
2. Click your avatar in the bottom-left corner, then select Settings
3. In the left sidebar, click Personal Access Tokens
4. Click Generate new token, give it a name (e.g., "Claude Desktop"), and click Create token
5. Copy the token immediately -- you won't be able to see it again

Step 2: Add to Claude Desktop

Open Claude Desktop. Go to Settings > Developer > Edit Config.

Replace the file contents with:

{
  "mcpServers": {
    "lightdash": {
      "command": "npx",
      "args": ["-y", "lightdash-mcp"],
      "env": {
        "LIGHTDASH_API_KEY": "your-token-here",
        "LIGHTDASH_API_URL": "https://your-lightdash-instance.com"
      }
    }
  }
}

• your-token-here -- the token from Step 1
• https://your-lightdash-instance.com -- your Lightdash URL (do NOT add /api/v1, the server adds it automatically)

If you already have other MCP servers, add just the "lightdash": { ... } block inside your existing mcpServers object.

Step 3: Restart and verify

Quit Claude Desktop completely (Cmd + Q), reopen it, and ask Claude:

Use the lightdash_ping tool to test the connection.

If Claude confirms the connection is working, you're all set.

Available Tools

| Tool | What It Does | |------|-------------| | lightdash_ping | Test that the connection to Lightdash is working | | list_projects | List all projects in your Lightdash organization | | list_spaces | List spaces within a project | | search_charts | Search for charts by name | | list_dashboards | List dashboards, optionally filtered by name | | list_explores | List available explores (data models) in a project | | get_chart | Get the full configuration of a saved chart | | get_chart_results | Run a saved chart and get its data | | get_explore | Get the schema of an explore (dimensions and metrics) | | run_raw_query | Run a custom query against an explore with filters and sorts | | run_chart_with_filters | Run a saved chart with optional filter overrides (deep-merged with chart's built-in filters) |

Troubleshooting

Check that Node.js v18+ is installed by running node --version in Terminal. npx requires Node.js. If you get "command not found", install Node.js from nodejs.org (choose the LTS version).

After installing, close and reopen Terminal, then restart Claude Desktop.

Your Personal Access Token is invalid or expired. Go to Lightdash Settings > Personal Access Tokens and create a new one. Then update LIGHTDASH_API_KEY in your Claude Desktop config with the new token.

Your LIGHTDASH_API_URL is wrong. Common mistakes:

• Adding /api/v1 at the end (don't -- the server adds this automatically)
• Using http:// instead of https://
• Typos in the URL

Alternative: Manual Install

If you prefer to clone the repository instead of using npx:

git clone https://github.com/timeleft-dev/lightdash-mcp.git
cd lightdash-mcp
npm install
npm run build

Then use this Claude Desktop config instead (replace /FULL/PATH/TO/ with the actual path):

{
  "mcpServers": {
    "lightdash": {
      "command": "node",
      "args": ["/FULL/PATH/TO/lightdash-mcp/build/index.js"],
      "env": {
        "LIGHTDASH_API_KEY": "your-token-here",
        "LIGHTDASH_API_URL": "https://your-lightdash-instance.com"
      }
    }
  }
}

Optional: Improve Claude's Output

You can add instructions to Claude Desktop's preferences so Claude formats Lightdash data nicely by default.

CSV Tables

Go to Claude Desktop Settings > General. In the "How would you like Claude to behave?" box, add:

When showing tabular data from Lightdash, format it as a CSV code block for easy copying.

Chart Artifacts

In the same preferences box, also add:

When I ask for charts or visualizations of Lightdash data, create them as artifacts using Recharts (a React charting library that Claude Desktop supports natively).

License

MIT