An AI Smart Contract Auditing Platform to detect for potential vulnerabilities in Solidity code.
This project utilises Data Flow Graph (DFG) to analyse the Smart Contracts, specifically Solidity as other programming languages does not have their vulnerable codes public. The Smart Contracts will be converted to Abstract Syntax Tree (AST) initially for it to be passed on to the model. The model selected here is Graph Neural Network (GNN) and it will detect for similar patterns depending on the supervised learning class. In the system itself, there are 2 main function. The first function which allow the users to select a solidity file to be inserted for vulnerability detection. Another function allow users to submit a Smart Contract that contain vulnerability for future AI training.