Infinite-Zero is the ultimate open-source hardware hacking ecosystem. It merges the deep RF of HackRF One, the advanced RFID of Proxmark3, and the pocket-sized agility of Flipper Zero and M5Stack. It acts as an AI-powered bridge between low-level embedded hardware (M5Stack Cardputer / ESP32-S3, HackRF One) and the Gemini API, transforming raw machine noise and RF signals into plain English insights and actionable execution.
The framework is split into two physical domains:
- The Host (Bridge): Runs the Python bridge script, interfaces with the Gemini API, and acts as the reasoning engine. Can be run on a laptop, Raspberry Pi, or Termux.
- The Edge (Hardware): Interacts with the physical environment. Captures RF, Wi-Fi, BLE, and RFID data, and executes injected payloads.
bridge_host.py: The primary LLM tether and communication orchestrator.generate_payload.py: Dynamically generates BadUSB/Duckyscript logic based on LLM analysis.system_prompt.txt: The system instructions bounding the AI's behavior as an onboard cybersecurity copilot.
The framework is split into two physical domains:
- The Host (Bridge): Runs the Python bridge script, interfaces with the Gemini API, and acts as the reasoning engine. Can be run on a laptop, Raspberry Pi, or Termux.
- The Edge (Hardware): Interacts with the physical environment. Captures RF, Wi-Fi, BLE, and RFID data, and executes injected payloads.
bridge_host.py: The primary LLM tether and communication orchestrator.generate_payload.py: Dynamically generates BadUSB/DuckyScript logic based on LLM analysis.system_prompt.txt: The system instructions bounding the AI's behavior as an onboard RF/security expert.wifi_scanner.py: Captures 802.11 network packets and handshakes.ble_scanner.py: Tracks and maps Bluetooth Low Energy beacons.bad_kb.py: Executes automated keystroke injection via HID emulation.
- Python 3.10+
- PlatformIO (for flashing Edge hardware)
- Google Gemini API Key
It completely changes the workflow of hardware hacking by using Large Language Models to abstract the complexity. Instead of capturing a raw, chaotic hex dump from an RFID tag or an SDR waterfall and manually decoding it, Infinite-Zer0 automates the translation. The AI acts as an onboard RF and cybersecurity expert, identifying protocols, explaining how they work, and dynamically generating BadUSB DuckyScripts on the fly.
The steepest barrier to entry in signal intelligence and hardware penetration testing is the massive learning curve. Interpreting raw sub-GHz frequencies or manually writing bitwise logic for protocol decoding requires deep, specialized knowledge. This framework solves that by turning raw machine noise into readable insights and automated feedback loops.
Like any penetration testing infrastructure, it is a dual-use toolkit. While bounded by system prompts against malicious use, the framework itself contains modules for BadUSB keystroke injection (bad_kb.py), Wi-Fi scanning (wifi_scanner.py), and Bluetooth Low Energy tracking (ble_scanner.py). In the wrong hands, this exact hardware stack can be used for unauthorized physical access, credential theft, and network intrusion.
1. Clone the Repository
git clone [https://github.com/credkellar-boop/Infinite-Zer0.git](https://github.com/credkellar-boop/Infinite-Zer0.git)
cd Infinite-Zer0