Automates the process of sending queries to a GUI-based LLM interface (e.g., Grok) using PyAutoGUI. This tool simulates human interaction by detecting the input box on-screen and typing in queries from a provided list.
- Visual query box detection via screenshot matching
- Simulated typing and submission of queries
- Batch querying support with average timing logs (WIP)
- Debug screenshots for troubleshooting [optional]
- Python 3.x
- PyAutoGUI
- OpenCV support (for confidence matching):
pip install opencv-python
- Capture a screenshot of the query input box and name it "grok.png" or "grok_1.png"
- Ensure the target LLM window is visible on the screen
- Run
py main.py
- Make sure display scaling is set to 100%
- Adjust the confidence level in locateOnScreen() if matching fails.