HacxGPT is an experimental adversarial AI framework designed for researchers investigating autonomous reasoning, prompt security, and model resilience. It provides a controlled environment to simulate unrestricted LLM behavior, benefiting security analysts and AI researchers by offering a platform for testing model vulnerabilities and adaptive prompt strategies.

How It Works

This framework is inspired by WormGPT's architecture, featuring an adaptive prompt layer that dynamically adjusts context and control logic. It integrates with external LLM providers like OpenRouter and DeepSeek via modular API connectors, allowing for the emulation of advanced, potentially unrestricted AI behaviors within defined ethical and technical boundaries. The system is optimized for safe experimentation and analysis of LLM responses.

Quick Start & Requirements

• Primary install/run command: pip install -r requirements.txt followed by python HacxGPT.py.
• Prerequisites: Python 3.10+, an API key from a supported provider (OpenRouter or DeepSeek), and a Windows operating system.
• Setup: Requires cloning the repository, installing dependencies, and running an installer script (install.bat). API key is configured on first launch or via HacxGPT.py.
• Links: OpenRouter, DeepSeek

Highlighted Details

• Autonomous Reasoning Framework for simulating near-unrestricted LLM behavior.
• Adaptive Prompt Layer for dynamic context and control logic adjustment.
• Cross-Provider Compatibility supporting multiple LLM APIs.
• Lightweight Windows Command-Line Interface (CLI) for experimentation.

Maintenance & Community

No specific details regarding contributors, community channels (like Discord/Slack), or roadmaps are provided in the README.

Licensing & Compatibility

Distributed under the MIT LICENSE. This license permits commercial use and integration into closed-source projects, with standard attribution requirements.

Limitations & Caveats

The project is described as "highly experimental" and a "demonstration framework." The full, fine-tuned HacxGPT model remains private and is not part of this open-source release. The framework is explicitly optimized and designed for Windows systems, lacking cross-platform support.