Prometheus

You set up this MCP server to bridge between Claude and your Prometheus instance, exposing metrics data and metadata through the Model Context Protocol. It lets you list available Prometheus metrics, read detailed metric information, and access current statistics, all via MCP-compatible interactions.

How to use

Use this MCP server to query Prometheus metrics from your preferred MCP client. You can list all available metrics, view each metric’s name and description, read Prometheus metadata (such as help text and type), and retrieve current statistics like count, minimum, and maximum values. Basic authentication is supported for secured Prometheus instances.

How to install

Prerequisites you need before installing this MCP server are Node.js and npm. Ensure you have JavaScript/TypeScript tooling available and a running Prometheus instance reachable from the server.

Install dependencies and build the server locally, then configure it for use with your MCP client.

npm install

npm run build

npm run watch

Configuration and MCP setup

Configure the MCP server to point at your Prometheus instance and, if needed, provide basic authentication credentials. The following example shows how you would register this MCP server for Claude Desktop. Adjust the path to the built index.js and the Prometheus URL to your environment.

{
  "mcpServers": {
    "mcp-server-prometheus": {
      "command": "/path/to/mcp-server-prometheus/build/index.js",
      "env": {
        "PROMETHEUS_URL": "http://your-prometheus-instance:9090"
      }
    }
  }
}

Additional sections

Configuration details include the base Prometheus URL and optional authentication credentials. The server exposes HTTP and stdio endpoints to communicate with MCP clients. Environment variables shown here are what you would pass when starting the server to enable connectivity and authentication as needed.

Security notes: If your Prometheus instance requires authentication, provide the username and password as additional environment variables and ensure the MCP client has access to these credentials in a secure manner.

Examples and tools

The server supports listing metrics and retrieving detailed metric information, including metadata and current statistics. Use these capabilities through your MCP client to build dashboards, automate monitoring checks, or integrate Prometheus data into broader workflows.

Available tools

List metrics

List all available Prometheus metrics with their descriptions.

Read metric

Read detailed metric information including metadata, help text, type, and current statistics (count, min, max).