Analysis updated 2026-05-18
Automatically time a throw at a creature using basic screen-watching mode.
Use YOLO-based detection to spot and aim at creatures in the game window.
Run the cruise mode to patrol a map automatically, pausing during battles.
Label new screenshots and train custom YOLO models for creatures not already covered.
| makapic/rocopilot | frayude/throttnux | linke-ai/hermes-agent-team | |
|---|---|---|---|
| Stars | 63 | 63 | 63 |
| Language | Python | Python | Python |
| Setup difficulty | hard | moderate | moderate |
| Complexity | 4/5 | 3/5 | 3/5 |
| Audience | general | ops devops | developer |
Figures from each repo's GitHub metadata at analysis time.
Requires an NVIDIA GPU, admin rights to install a kernel driver, and a system reboot before first use.
RocoPilot is an automation assistant for a Chinese online game called Luoke Kingdom: World. It watches your screen using computer vision, identifies game elements, and then moves your mouse and keyboard automatically in response. The README is written in Chinese. The tool has three operating modes. The first simply watches for a moment when a ball can be thrown at a creature and clicks at the right time. The second adds AI-powered creature detection: a YOLO vision model spots a creature in the game window, the tool aims the camera at it, and throws the ball. The third mode extends that with a cruise system that automatically walks a patrol route through the map searching for creatures, pausing whenever a battle starts and resuming when it ends. To bypass the game's anti-cheat system, the tool uses a kernel-level input driver called Interception. Rather than injecting code into the game process or changing game memory, it simulates a physical mouse and keyboard at the operating system level, which the anti-cheat cannot distinguish from a real player. The README is explicit that using this violates the game's terms of service and carries risk of account suspension. The vision side uses OpenCV template matching for recognizing menus and buttons, plus pre-trained YOLO models (provided in the repository) for detecting specific creature types. The project also includes scripts for labeling screenshots and training new models for additional creatures not covered by the included ones. The tool runs on Windows only and requires an NVIDIA GPU for the AI detection features. Setup involves installing the Interception driver with administrator rights, rebooting, and then running the Python installer. A pre-built Windows executable is also provided for users who do not want to set up a Python environment.
A vision-based automation bot for a Chinese mobile game that spots creatures on screen and controls the game via a kernel-level input driver.
Mainly Python. The stack also includes Python, OpenCV, YOLO.
The README does not state a specific license for this project.
Setup difficulty is rated hard, with roughly 1h+ to a first successful run.
Mainly general.
This repo across BitVibe Labs
Verify against the repo before relying on details.