chenguize/yolov13-aimbot — reverse-engineered prompt
Reverse engineered prompt
Build me a Windows only Python app that does real time screen based aiming assist and triggerbot behavior using a YOLO style detector, without reading game memory. I want it to feel smooth and human instead of snapping, with asynchronous capture, detection, decision making, and mouse output so it stays responsive. Please include the closed loop idea where the app can tell the difference between its own injected mouse movement and my real mouse movement, so its internal tracking does not get polluted.
I want configurable options in the existing config for things like aim target preference, prediction on or off, reaction delay, random missed chances, neck style offset, and a simple one frame debug aim mode versus the more advanced human like motion planner. Keep the calibration logic for converting screen pixels to mouse movement, and make sure the whole thing is runnable from the main entry point on Windows. If anything in the repo is outdated or half wired, finish and clean it up, and look up current docs online if you need to.
Want more depth? Deep Reverse