Umami MCP Server

Umami MCP Server is an MCP server that connects self-hosted Umami Analytics or Umami Cloud instances to AI assistants and development environments. It allows website owners, software engineers, data analysts, and marketers to query real-time traffic, aggregated performance stats, and individual session activity directly within an AI chat interface. By integrating through the Model Context Protocol, users can inspect pageviews, unique visitors, bounce rates, visit durations, and granular visitor demographics like browsers, operating systems, countries, and devices without opening the Umami dashboard. The server also exposes tools for tracking live active users and breaking down chronological session actions to review user navigation paths. It supports both local command execution via Go binaries and Docker, as well as remote streamable HTTP deployments using preconfigured endpoints.

Category: Data & Analytics

Tags: analytics, dashboard, metrics, umami, web-traffic

Visit Umami MCP Server

How to install and configure Umami MCP Server

1. Choose between local installation or the remote HTTP endpoint. For local use, install via Go with go install github.com/Macawls/umami-mcp-server@latest, pull the Docker image ghcr.io/macawls/umami-mcp-server, or download a release binary. 2. Open your client configuration file, such as Claude Desktop (claude_desktop_config.json) or Cursor (.cursor/mcp.json). 3. Add the server entry under mcpServers. For a local binary setup, define command with the path to umami-mcp-server and supply UMAMI_URL, UMAMI_USERNAME, and UMAMI_PASSWORD (or UMAMI_API_KEY for Umami Cloud) under env. 4. Alternatively, use the remote endpoint by specifying "type": "http" and "url": "https://umami-mcp.macawls.dev/mcp", passing credentials via X-Umami-* headers. 5. Save the configuration and restart your client.

What you can do with Umami MCP Server

• Reviewing overall website performance metrics, pageviews, and bounce rates across custom time windows directly inside Claude Desktop. - Identifying high-performing content and decaying blog posts to plan editorial updates and track marketing campaigns using UTM parameters. - Inspecting chronological user navigation sequences and session histories to analyze how specific visitors move through site funnels. - Monitoring current active visitor volume in real time to detect immediate server traffic spikes or sudden service outages.

Key facts

• Open Source
• https://github.com/Macawls/umami-mcp-server
• Data & Analytics, Monitoring & Observability
• analytics, dashboard, metrics, umami, web-traffic

Part of MCP Servers

How do I install Umami MCP Server?

You can install it locally using Go by running go install github.com/Macawls/umami-mcp-server@latest, run it via Docker with ghcr.io/macawls/umami-mcp-server, download a prebuilt binary from GitHub, or use the hosted remote HTTP endpoint at https://umami-mcp.macawls.dev/mcp without installing software.

What can Umami MCP Server do?

It provides tools to list websites, retrieve aggregated traffic statistics, group pageviews over time, break down metrics by browser or country, check active visitor counts in real time, list individual user sessions, and retrieve detailed event timelines for specific visits.

Which MCP clients work with Umami MCP Server?

It works with any MCP-compatible environment that supports stdio or HTTP transports, including Claude Desktop, Cursor, VS Code with GitHub Copilot, Claude Code, Windsurf, Zed, OpenCode, and Smithery.

Does Umami MCP Server work with Umami Cloud?

Yes. While self-hosted instances typically use a username and password, Umami Cloud requires an API key. You set the instance URL to https://api.umami.is and provide your UMAMI_API_KEY via environment variables, config.yaml, or HTTP headers.

Is Umami MCP Server open source?

Yes, Umami MCP Server is an open source project available on GitHub under the repository Macawls/umami-mcp-server.