This tool utilizes YOLOv5 for high-speed, real-time detection of humanoid characters. While the AI logic is based on the RootKit-Org framework, this project is optimized for flexibility:
- Hybrid Input Support: You can run the bot entirely via Software (Windows API) for a quick start, or use an Arduino Leonardo for professional-grade Human Interface Device (HID) hardware emulation.
- Security Focused: The hardware interface is designed to provide the safest possible mouse movement, making it look like a genuine physical device to any Anti-Cheat.
- Performance: Optimized for low latency, whether you are using CUDA-powered NVIDIA cards or modern AMD hardware.
Modern anti-cheat systems often block virtual mouse inputs. This project demonstrates how software-based restrictions can be bypassed using an Arduino Hardware Bridge:
- HID Proxy: An Arduino Leonardo acts as a physical mouse.
- Hardware Signals: Mouse commands are sent as genuine USB signals, making software-level detection nearly impossible.
- Awareness: The goal is to raise awareness among developers regarding these hardware-based vulnerabilities.
⚠ Important Note: Use at your own risk. If you get caught, you’ve been warned! I assume no liability for any consequences or game bans. Use this knowledge responsibly!
If you have questions, feel free to add me on Discord:
👤 Discord: Foxi7
- Standard Emulation: Uses win32api.
- Hardware Bridge (Arduino): Uses an Arduino Leonardo for genuine hardware signals (Safest Method).
- NVIDIA: CUDA Cores (Maximum Speed).
- AMD / DirectML: GPU acceleration for AMD graphics cards.
- CPU: Runs on any machine (slower).
- GPU (NVIDIA): GTX 10-series or newer & NVIDIA CUDA Toolkit 11.8 (Recommended for speed).
- GPU (AMD): DirectX 12 compatible.
- Input Method (Choose one):
- Software Emulation: Uses standard Windows API. No extra hardware needed—works instantly.
- Hardware Bridge: Supports ATmega32U4-based boards (e.g., Leonardo, Pro Micro). This provides native HID mouse signals for maximum bypass security.
- Additional Software: Arduino IDE (Only if you use a hardware bridge).
-
Download: Click the green Download button at the top or download the source code from the Latest Release and extract the archive to a folder 🗂️.
-
Python: Install Python 3.11.x (Important: Check "Add Python to PATH" during installation!) 🐍.
-
Hardware Setup (Optional - for Arduino Users):
- Connect: Plug your ATmega32U4-based board (e.g., Leonardo, Pro Micro) into your PC via USB.
- Identify Port: Open the Windows Device Manager, expand the Ports (COM & LPT) section, and note the COM port assigned to your board (e.g.,
COM3). - Open Sketch: Open the file
Arduino_Mouse_HID.ino(located in the project'sArduinofolder) using the Arduino IDE. - Select Board & Port: - Go to Tools > Board and select Arduino Leonardo (Select this even if you are using a Pro Micro, as it uses the same chip!).
- Go to Tools > Port and select the exact COM port you identified in step 2.
- Flash: Click the Upload arrow (top left) to flash the code to your hardware. Note: You will need to enter this same COM port later in the Bot's S-Menu.
-
Installation & Terminal Setup:
-
IMPORTANT: Before running the commands below, open PowerShell or CMD and navigate to the extracted project folder:
- Option A (Easy): Type
cd(with a space), drag the project folder into the terminal window, and press Enter. - Option B (Manual): Navigate via path, e.g.:
cd "C:\Path\To\Your\Project".
- Option A (Easy): Type
-
Nvidia GPU Users:
pip install torch==2.2.2 torchvision==0.17.2 torchaudio==2.2.2 --index-url [https://download.pytorch.org/whl/cu118](https://download.pytorch.org/whl/cu118) pip install onnxruntime-gpu==1.17.1 pip install cupy-cuda11x
-
AMD or CPU Users:
pip install torch torchvision torchaudio
-
Final Step (Required for all):
pip install -r requirements.txt
-
-
Game Preparation: Set your game to Windowed or Borderless Window mode. 🖥️
-
Terminal Navigation: Open PowerShell or CMD and navigate to the project folder:
- Option A (Easy): Type
cd(with a space), drag your folder into the terminal, and press Enter. - Option B (Manual): Type the full path, e.g.,
cd "C:\Users\Name\Documents\Foxbot-AI".
- Option A (Easy): Type
-
Start: Run the script with:
python main.py -
The S-Menu 🛠️: Press 'S' for the interactive setup:
- Navigation: Type your value and press ENTER to confirm, or simply press ENTER to skip a setting and keep its default.
- Arduino Users: When prompted, enable Arduino mode (
y) and enter your COM Port (e.g.,COM3orCOM9). - Engine Selection: Choose your hardware engine when prompted (
1= CPU,2= AMD,3= NVIDIA).
-
Final Launch 🚀: Press ENTER and choose your game window to arm the bot, then switch to your game.
Tip
You can skip the S-Menu entirely by manually editing the config.py file with any text editor to save your preferred settings without the script.
- [CAPS] 🎯: Master Switch (Toggles the Aimbot ON/OFF).
- [PAGEDOWN] 🔄: Mode Toggle (Always-On vs. Hold-to-Aim).
- [INSERT] 🔫: Triggerbot Switch (Toggles Auto-Fire ON/OFF).
- [END] 💣: Exit (Closes the script immediately).
| Feature | Variable | Default | Description |
|---|---|---|---|
| 🏎️ Performance | visuals |
False |
Preview window with AI boxes (Keep False for max FPS) |
onnxChoice |
1 |
Device: 1=CPU, 2=AMD, 3=NVIDIA | |
| 🔌 Hardware | use_arduino |
False |
True for Leonardo HID / False for win32api |
arduino_port |
'COM?' |
Needs to be set to your COM Port (e.g. 'COM7') | |
| 🎯 Aiming | MovementAmp |
0.4 |
Speed/Smoothing. Dependent on In-Game Sense! |
confidence |
0.4 |
Detection threshold (Lower = more aggressive detection) | |
centerOfScreen |
True |
Prioritizes targets closest to your crosshair | |
| 🧠 Targeting | headshot_mode |
False |
Toggles between Head and Body aim |
headshot_offset |
0.38 |
Height adjustment (Values vary by game/character size: 0.38 = Head, 0.2 = Chest) | |
| 🔫 Triggerbot | triggerbot_enabled |
False |
Independent auto-fire status |
trigger_radius |
15 |
NEW: Maximum pixel distance around target point for instant firing | |
| ⌨️ Controls | hotkeyAimbot |
'CAPS' |
Toggle key to activate/deactivate the bot |
hotkeyRMB |
'PAGEDOWN' |
Switch for "Hold-to-Aim" mode | |
hotkeyDelay |
0.25 |
Delay in seconds before Aim kicks in (RMB Mode) | |
hotkeyTrigger |
'INSERT' |
NEW: Toggle key to activate/deactivate the Triggerbot | |
QuitKey |
'END' |
Emergency stop key for the script |
Tip
Performance Boost: The option is visuals = False by default to ensure the lowest possible input lag. Only enable it if you want to debug the AI detection visually.
Tip
Accuracy & Sensitivity: If the bot "shakes" or overshoots, lower your MovementAmp. Note that your In-Game Sensitivity directly affects this: Higher in-game sense requires a lower MovementAmp to stay smooth. A value between 0.3 and 0.5 is usually the sweet spot.
Tip
Finding your COM Port: If you are using an Arduino, open the Windows Device Manager, look under Ports (COM & LPT), and find the number assigned to your "Arduino Leonardo". Enter this in the config.py (e.g., 'COM7').
Features marked with [x] are already integrated and working:
- Hybrid Input: Support for both Arduino Hardware and Software Mouse
- Cross-Platform GPU: Acceleration via CUDA (NVIDIA) and DirectML (AMD)
- S-Menu Configuration: Change settings like Amp and Confidence on the fly
- Adjustable Smoothing: Integrated movement amplification for better control
- Triggerbot: Auto-fire when a target is locked 🔫 (New: Instant execution via custom pixel radius)
- Custom Game Models: Dedicated AI weights for different games
- TensorRT Support: Conversion to
.enginefor maximum NVIDIA performance 🏎️ - Bezier Curves: Researching human-like mouse paths (Bezier/Splines)
Caution
Regarding Custom Models: Use of game-specific models increases the risk of detection by Anti-Cheat systems. These features are intended for educational and offline research purposes only. I do not take responsibility for any bans or consequences.
- Basis: RootKit-Org (Core logic, config & selection).
- AI Engine: YOLOv5 by Ultralytics.
This project is licensed under the GNU General Public License v3.0.
See the LICENSE file for more details. Based on the work of RootKit-Org.
Have fun with the project! 🎉👾