Metamorphic Code Engine using LLMs

A research prototype designed to generate metamorphic C code variants using Large Language Models (LLMs). This tool provides a graphical user interface to manage the entire lifecycle of code mutation, from prompt engineering and generation to compilation, analysis, and evaluation.

It supports both local models (via LM Studio) and cloud-based APIs (OpenAI, Grok) and was developed as part of a master's thesis on automated malware obfuscation.

✨ Key Features

🏗️ Architecture Overview

📋 Prerequisites

📦 Installation

⚙️ Configuration

On first run, the application generates a config.ini file. Edit it to add API keys and configure local models.

[API_KEYS]
GPT = YOUR_OPENAI_API_KEY_HERE
GEMINI = YOUR_GOOGLE_API_KEY_HERE
GROK = YOUR_API_KEY_HERE
COPILOT = YOUR_API_KEY_HERE
VIRUSTOTAL = YOUR_VIRUSTOTAL_API_KEY_HERE
[LOCAL_MODELS]
model1 = mistralai/mistral-small-3.2
model2 = qwen/qwen3-coder-30b

🚀 Usage

python main.py

🧠 Model Selection

Select Model Interface

🖥️ Main Interface

Tabs allow editing of instructions, code, and context. The sidebar provides access to mutation control, compilation, statistics, and VirusTotal analysis.

Main Interface

📊 Statistics Collector

Statistics Report

🦠 VirusTotal Analysis

VirusTotal Report