Business Compliance Detection of Smart Contracts in Electricity and Carbon Trading Scenarios

Abstract

Business compliance in smart contracts for blockchain-based electricity and carbon trading (B-ECT) remains unexplored. We propose an automated Business Compliance Detection tool for smart contracts (BCDetection ) in B-ECT to address this gap. Our innovation encompasses the creation of a benchmark dataset containing both compliant and non-compliant smart contracts, coupled with the deployment of Agent-based Large Language Models (LLMs) to align smart contract codes with prevailing business regulations. The BCDetection tool employs a structured agent for compliance verification, including pre-judgment, feature extraction, fine-grained feature alignment, and consistency judgment. A case study demonstrates its effectiveness. As the field evolves, our approach shows promise for enhancing security and compliance.