PROMFUZZ is an automated and scalable system that leverages LLM-driven, bug-oriented composite analysis to effectively identify functional bugs in smart contracts.

Its core design is based on three key insights:

1. LLM-Driven Multi-Perspective Analysis
2. Dual-Stage Invariant Checker Generation
3. Bug-Oriented Fuzzing Engine

• Docker
• Python 3.7+
• Pip

• See: Build Instructions

• This step must be completed first since you need the container ID.

pip install -r requirements.txt

• Linux / macOS:

export OPENAI_API_KEY="your_api_key"

• Windows PowerShell:

$env:OPENAI_API_KEY = "your_api_key"

• Get your API key from: https://platform.openai.com/account/api-keys

• To run PROMFUZZ, use the following command:

python promfuzz.py \
    --input=<SOLIDITY_FILE_PATH> \
    --containerid=<ENGINE_CONTAINER_ID> \
    --enginetimeout=<ENGINE_TIMEOUT_IN_SECONDS>

• Example:

python promfuzz.py \
    --input=MyContract.sol \
    --containerid=ab12cd34ef56 \
    --enginetimeout=180

If you use PROMFUZZ in your research, please cite the following paper:

@inproceedings{promfuzz,  
  author = {Xingshuang Lin and Qinge Xie and Binbin Zhao and Yuan Tian and Saman Zonouz and Na Ruan and Jiliang Li and Raheem Beyah and Shouling Ji}, 
  title = {PROMFUZZ: Leveraging LLM-Driven and Bug-Oriented Composite Analysis for Detecting Functional Bugs in Smart Contracts},  
  booktitle = { {IEEE/ACM} International Conference on Automated Software Engineering ({ASE}) },  
  year = {2025},  
}