prometheus-mcp-server

MCP.Pizza Chef: prometheus

This MCP server connects your AI assistant to a running Prometheus monitoring system, enabling it to query metrics, analyze performance, and suggest improvements. You can ask your AI to investigate slow queries, review server health, and generate Prometheus query language commands. It requires running the server via Docker and connecting it to your Prometheus instance. Compatible with MCP-enabled clients like Claude Desktop, Cursor, or ChatGPT with MCP support.

Data

Other

Use This MCP server To

Investigate why Prometheus queries are slow Generate and execute Prometheus queries via AI List and analyze Prometheus metrics and labels Get suggestions for Prometheus recording rules for SLOs Summarize Prometheus metric and label naming best practices Review health and configuration of a Prometheus server Search official Prometheus documentation from AI

README

Prometheus MCP Server

About

This is an MCP server to allow LLMs to interact with a running Prometheus instance via the API to do things like generate and execute promql queries, list and analyze metrics, etc.

Demos and Examples

Asking Claude to Investigate Slow Queries

The prompt used was:

querying my metrics is slow, can you help me figure out why?

prometheus-mcp-server FAQ

Can I use this to analyze and query Prometheus metrics?

Yes — this server lets your AI generate and run Prometheus queries, list metrics, and analyze data from a Prometheus instance.

Can I use this to get suggestions for Prometheus recording rules?

Yes — it can investigate metrics and suggest recording rules for service level objectives (SLOs).

Which apps can use this MCP server?

It is designed to work with any MCP-compatible client that supports server connections, such as Claude Desktop, Cursor, or ChatGPT with MCP support.

Do I need an API key or account to use this?

No special API key is needed, but you do need access to a running Prometheus server instance.

How hard is it to set up?

Setup requires developer skills since it involves running a Docker container and configuring connections to Prometheus.

Does it support multiple AI models?

Yes — you can use it with OpenAI, Anthropic Claude, and Google Gemini models through MCP clients.