Microsoft Clarity MCP Server: Tools, Setup and Security

The Microsoft Clarity MCP server lets an AI app query your Clarity analytics, list session recordings and search Clarity's documentation. You can ask questions like "What are the most used browsers in my Clarity project?" in plain English. Microsoft publishes it as an open-source npm package in its microsoft GitHub organization.

Checked against: Microsoft Clarity MCP server on GitHub on 28 October 2026. eSEOspace did not install or test this server.

If you want behavior data next to GA4 numbers, also see the Google Analytics MCP server.

At a glance

What it can do

The README documents three tools:

• query-analytics-dashboard: "Retrieves analytics data and metrics from your project's dashboard using a simplified natural language search query." Example: "How many Clarity sessions did we get from Egypt in the past 3 days?"
• list-session-recordings: lists recordings filtered by URL, device type, browser, OS, country, city and more. Example: "Get Clarity recordings where users encountered JavaScript errors."
• query-documentation-resources: pulls snippets from Clarity's documentation for setup and troubleshooting questions.

The README's configuration section also mentions passing a token to a get-clarity-data tool, which is not in the tool list above. Check the tool list in your client after you connect.

The underlying Clarity Data Export API supports dimensions such as browser, device, country, source, medium, campaign, channel and URL. Its metrics include scroll depth, engagement time, dead clicks, rage clicks and script errors.

How to connect it

First, create an API token in Clarity. The README says: go to your Clarity project, select Settings, then Data Export, then Generate new API token. Microsoft's docs note that only project admins can manage tokens.

Then add the server to your MCP client. The README's generic config:

{
  "mcpServers": {
    "@microsoft/clarity-mcp-server": {
      "command": "npx",
      "args": [
        "@microsoft/clarity-mcp-server",
        "--clarity_api_token=your-api-token-here"
      ]
    }
  }
}

For Claude Desktop, the README says to add this block to claude_desktop_config.json and restart. It also lists a Claude Desktop extension called Microsoft Clarity, and a one-click install button for VS Code.

Data limits

Microsoft's Data Export API docs list these limits:

• A maximum of 10 API requests per project per day.
• Data from the previous 1 to 3 days only.
• Up to three dimensions per request.
• Responses capped at 1,000 rows, with no pagination.

The README does not say how the server maps to these limits, so plan short, focused questions.

Security notes

• The token sits in your config. The README's example puts the token in the command arguments. Treat that file as sensitive; see MCP secret leakage.
• Rotate tokens when people leave. Microsoft advises replacing tokens promptly if a user with access is removed from the project.
• Recordings can hold personal data. Session recordings show real user behavior. Check that sending summaries to your AI provider fits your privacy notice. Clarity's privacy policy is linked from the README.
• Read-only tools. None of the documented tools change your Clarity setup, which lowers the risk.

The MCP security hub covers wider checks.

Alternatives

• Google Analytics MCP server: Google's local server for GA4 reports.
• Amplitude MCP server: hosted product analytics with read and write tools and session replay access.

Next step

Behavior data helps explain why pages convert or not. Pair it with your SEO and analytics data using the ideas in MCP for marketing teams, or ask about our Claude Code consulting.

Frequently asked questions