🚀 ⚡️ k6-mcp-server
About
A Model Context Protocol (MCP) server implementation for running k6 load tests. It integrates the k6 load testing tool into MCP clients (Claude Desktop, Cursor, Windsurf, etc.), enabling users to execute and manage k6 tests through natural language prompts.
Details
- Author
- MCP-Mirror
- Downloads
- 293
- Categories
- Other
Jump to
- Simple integration with Model Context Protocol framework
- Support for custom test durations and virtual users (VUs)
- Easy-to-use API for running k6 load tests
- Configurable through environment variables
- Real-time test execution output
Setting up with Highlight
This MCP is not yet compatible with Highlight’s one-click setup. However, you can still use it with Highlight by following these steps:
- Download and install Highlight from highlightai.com/download
- Navigate to the plugins tab and select "Add Custom Plugin"
-
Configure the plugin with the settings below
Plugin Name
🚀 ⚡️ k6-mcp-serverCommand (node, npx, python, etc.)Please refer to the README for specific instructions on how to obtain API keys or other required environment variables.
- Enable "Start Automatically" if you want the plugin to start when Highlight launches
From the repository
Install Python 3.12+, k6, and uv. Clone the repository and run uv pip install -r requirements.txt. Optionally set the K6_BIN environment variable. In your MCP client, configure the server with the command /path/to/bin/uv and args ["--directory", "/path/to/k6-mcp-server", "run", "k6_server.py"]. Then ask the LLM to run a k6 test (e.g., "run k6 test for hello.js").
Claude Desktop / Cursor
Paste into your MCP client config file to install this server.
{
"mcpServers": {
"\ud83d\ude80 \u26a1\ufe0f k6-mcp-server": {
"QAInsights_k6-mcp-server": {
"command": "uv",
"args": [
"pip",
"install",
"-r",
"requirements.txt"
]
}
}
}
}McpServers
{
"QAInsights_k6-mcp-server": {
"command": "uv",
"args": [
"pip",
"install",
"-r",
"requirements.txt"
]
}
}🚀 ⚡️ k6-mcp-server
A Model Context Protocol (MCP) server implementation for running k6 load tests.
✨ Features
- Simple integration with Model Context Protocol framework
- Support for custom test durations and virtual users (VUs)
- Easy-to-use API for running k6 load tests
- Configurable through environment variables
- Real-time test execution output
🔧 Prerequisites
Before you begin, ensure you have the following installed:
- Python 3.12 or higher
- k6 load testing tool (Installation guide)
- uv package manager (Installation guide)
📦 Installation
1. Clone the repository:
git clone https://github.com/qainsights/k6-mcp-server.git
2. Install the required dependencies:
uv pip install -r requirements.txt
3. Set up environment variables (optional):
Create a .env file in the project root:
K6_BIN=/path/to/k6 # Optional: defaults to 'k6' in system PATH
🚀 Getting Started
1. Create a k6 test script (e.g., test.js):
import http from "k6/http";
import { sleep } from "k6";
export default function () {
http.get("http://test.k6.io");
sleep(1);
}
2. Configure the MCP server using the below specs in your favorite MCP client (Claude Desktop, Cursor, Windsurf and more):
{
"mcpServers": {
"k6": {
"command": "/path/to/bin/uv",
"args": [
"--directory",
"/path/to/k6-mcp-server",
"run",
"k6_server.py"
]
}
}
}
3. Now ask the LLM to run the test e.g.
run k6 test for hello.js. The k6 mcp server will leverage either one of the below tools to start the test.
- execute_k6_test: Run a test with default options (30s duration, 10 VUs)
- execute_k6_test_with_options: Run a test with custom duration and VUs
📝 API Reference
Execute K6 Test
execute_k6_test(
script_file: str,
duration: str = "30s", # Optional
vus: int = 10 # Optional
)
Execute K6 Test with Custom Options
execute_k6_test_with_options(
script_file: str,
duration: str,
vus: int
)
✨ Use cases
- LLM powered results analysis
- Effective debugging of load tests
🤝 Contributing
Contributions are welcome! Please feel free to submit a Pull Request.
📄 License
This project is licensed under the MIT License - see the LICENSE file for details.
Sign in to leave a review
Use Google, GitHub, or an email account so ratings stay tied to real people.
No reviews posted yet.
