mcp-server-prometheus
The mcp-server-prometheus is a TypeScript-based server that interfaces with Prometheus to provide access to metrics and data through the Model Context Protocol. It offers features like detailed metric metadata, support for JSON structured data access, and capabilities to list and access Prometheus metrics.
What is the purpose of the mcp-server-prometheus?
The mcp-server-prometheus serves as a bridge between Claude AI and Prometheus, allowing for interaction and data retrieval from Prometheus metrics.
How do I configure the server for my Prometheus instance?
You need to set the PROMETHEUS_URL environment variable to the base URL of your Prometheus instance. Optionally, set PROMETHEUS_USERNAME and PROMETHEUS_PASSWORD for basic authentication.
What kind of data can I access with this server?
You can access Prometheus metric schemas, detailed metadata, and current statistical data such as count, min, and max values.
How can I debug the server?
Use the MCP Inspector tool for debugging, which provides a URL to access debugging tools in your browser.
What is the API structure for accessing metrics?
The server exposes metrics through URIs structured as http://your-prometheus-instance:9090/metrics/{metric_name}, returning JSON data.