feather_wand_agent

Feather Wand Agent is a unified AI-powered toolkit that orchestrates multiple performance testing tools—JMeter, k6, Gatling, and Locust—through a single conversational interface. You can ask the agent to run a JMeter test plan, execute a k6 script, run a Locust load test, or launch a Gatling simulation, and the agent will translate your natural language request into the appropriate tool commands and configurations. The agent exposes a web interface you can access after starting it (for example at http://localhost:8000) where you can interact with the chat to initiate tests, adjust parameters, and view results and metrics as tests run. The included utilities and environment-based configuration allow you to tailor the runner binaries, target hosts, and runtime parameters to fit your environment. This makes it suitable for performance testing in CI pipelines, local development, or monitoring scenarios where you want a conversational workflow to control tests across multiple tooling ecosystems.

Prerequisites:

• Python 3.8+ installed on your system
• Git installed
• JMeter (for JMeter tests) installed and accessible in PATH
• k6 installed and accessible in PATH
• Maven or Gradle installed for Gatling tests
• Locust installed and accessible in PATH

Installation steps:

1. Clone the repository

git clone https://github.com/yourusername/feather_wand_agent.git
cd feather_wand_agent

2. Create and activate a Python virtual environment (optional but recommended)

python -m venv venv
# macOS/Linux
source venv/bin/activate
# Windows
venv\Scripts\activate

3. Install Python dependencies

pip install -r requirements.txt

4. Configure environment variables (copy example and edit as needed)

cp .env.example .env

5. Run the agent server

python -m multi_tool_agent

6. Access the web UI (default):

• Open http://localhost:8000 in your browser

Optional: If you prefer running via a package manager style entry, you can adapt to your environment by using the MCP config below.

Tips and notes:

• Ensure the corresponding tooling binaries (JMeter, k6, Locust, Gatling) are installed and available in your system PATH or configured via environment variables (JMETER_BIN, K6_BIN, LOCUST_BIN, GATLING_RUNNER).
• Place your Google API key in the environment (.env) as GOOGLE_API_KEY if you plan to use Gemini integrations from Google AI Studio.
• The agent supports environment-based customization for performance test parameters (e.g., LOCUST_USERS, LOCUST_RUNTIME, LOCUST_HOST). Tune these in .env to fit your target environment.
• When running tests in CI, consider using non-GUI modes for JMeter and headless Locust to improve speed and stability.
• If you encounter issues with starting the web interface, verify that port 8000 is not in use by another process or adjust the host/port configuration if your environment requires it.
• For Gatling, choose a runner (mvn or gradle) compatible with your project setup and ensure the project contains a valid Gatling simulation in the specified directory.
• Check unit tests under tests/ for quick validation of the agent utilities before integrating into larger workflows.