Automated Vulnerability Pre‐Detection for Smart Contracts: A Blockchain‐Based Data Sharing Platform With Incentivized Consensus
Fujun Wang, Ziniu Shen, Yunfang Chen · Concurrency and Computation Practice and Experience · 2025
ABSTRACT The proliferation of smart contracts has significantly fueled the growth and advancement of decentralized industries. Nonetheless, flaws in smart contract design and the infancy of their applications have resulted in a persistent rise in contractual vulnerabilities. Tools aimed at detecting vulnerabilities within smart contracts are hampered by low accuracy, stemming from a scarcity of standardized training data. Additionally, smart contract datasets are plagued by a myriad of challenges, including outdated information, inaccuracy, and an absence of universal standards, which have hindered their broad acceptance. To address these challenges, this paper introduces a blockchain‐based solution for a smart contract data sharing community. It devises a novel consensus‐based incentive mechanism to curate a high‐quality smart contract dataset and integrates a suite of automated detection tools to create an efficient pre‐detection methodology for vulnerabilities. Subsequently, the paper implements a smart contract data sharing platform. The experimental results demonstrate that mainstream detection tools fail to identify all vulnerabilities. Notably, detection rates for reentrancy, illegal delegate call, integer overflow, and lock of balance vulnerabilities exceeded 50%, with reentrancy detection reaching 88.7%. Furthermore, our framework identified 681 previously unmarked vulnerabilities. These findings robustly validate the feasibility and effectiveness of the proposed incentive scheme in enhancing vulnerability detection accuracy and data quality.