qainsights/locust-mcp-server

Overview

An MCP server that lets you run Locust load tests from AI coding assistants. It wraps Locust's distributed load testing engine behind the Model Context Protocol so tools like Claude can start, monitor, and stop tests programmatically.

Best for

Best for
Developers who want to kick off Locust tests from an AI assistant without switching tools

Use cases

• Trigger a load test from a chat prompt during code review
• Check test results and failure rates without leaving the IDE
• Automate performance regression checks in a CI pipeline

How to use

Install

uv pip install -r requirements.txt

Tools exposed

• run_locust

Tested with

Claude Desktop, Cursor, Windsurf

Example client config

{\n  "mcpServers": {\n    "locust": {\n      "command": "/Users/naveenkumar/.local/bin/uv",\n      "args": [\n        "--directory",\n        "/Users/naveenkumar/Gits/locust-mcp-server",\n        "run",\n        "locust_server.py"\n      ]\n    }\n  }\n}

Notes

An MCP server that lets you run Locust load tests from AI coding assistants. It wraps Locust’s distributed load testing engine behind the Model Context Protocol so tools like Claude can start, monitor, and stop tests programmatically.

12 stars on GitHub. Last updated 2025-04-06. Licensed MIT.

Use cases

• Trigger a load test from a chat prompt during code review
• Check test results and failure rates without leaving the IDE
• Automate performance regression checks in a CI pipeline

Pros

• Brings load testing into conversational workflows
• Lightweight Python implementation with few dependencies
• Open source with a clear, focused scope

Cons

• Very early stage with only 12 GitHub stars
• Limited documentation and community support
• Requires a running Locust environment to be useful

Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.