prometheus_mcp_server

Overview

This Python-based MCP server enables Large Language Models to retrieve and analyze metric data from Prometheus databases. It provides tools for fetching specific metrics, performing statistical analysis, searching metric usage patterns, and executing complex PromQL queries. The server supports custom time ranges for data analysis and is designed to be integrated with AI environments like Claude Desktop.

Try asking AI

After installing, here are 5 things you can ask your AI assistant:

• retrieve_metric_info

  Retrieve comprehensive metric information including names and descriptions from Prometheus

  Try:Show me all available metrics in PrometheusTry:What metrics are currently being tracked?
• fetch_metric_data

  Fetch and analyze specific metric data using metric names

  Try:Get CPU usage metrics for the last hourTry:Retrieve memory consumption metrics for my application
• analyze_metric_data

  Analyze metric data within custom time ranges

  Try:Analyze response time metrics for the last 24 hoursTry:Compare error rates between this week and last week
• execute_promql_query

  Execute advanced PromQL queries for in-depth data exploration

  Try:Run PromQL query to find top 5 slowest endpointsTry:Execute PromQL to get average request latency per minute

Note: Tool names were inferred from the capabilities section of the README as there is no explicit 'Tools' section listing the MCP functions. The actual tool names and signatures are not explicitly documented in the README.

Comparable tools

Installation

npx -y @smithery/cli install @CaesarYangs/prometheus_mcp_server --client claude

cd ./src/prometheus_mcp_server
python3 -m venv .venv
source .venv/bin/activate

pip install -r requirements.txt

{
    "mcpServers": {
        "prometheus": {
            "command": "uv",
            "args": [
                "--directory",
                "/path/to/prometheus_mcp_server",
                "run",
                "server.py"
            ],
            "env": {
                "PROMETHEUS_HOST": "http://localhost:9090"
            }
        }
    }
}

FAQ

What Prometheus operations are supported?

The server supports retrieving metric information, fetching and analyzing specific metrics, analyzing data within custom time ranges, and executing PromQL queries. Label filtering is marked as in development.

Can I use this with other MCP clients besides Claude Desktop?

Yes, the server can be used standalone with any MCP client. The README provides instructions for using it with Cursor environment and as a standalone server.

Compare prometheus_mcp_server with