一句话：提出以审计员为中心的静态分析框架 AllocScope，通过建模资金分配流语义自动发现智能合约中的分配操纵漏洞。
DeFi 合约中错误的资金分配逻辑会导致代币分配不公，而每个操作单独看都合法，现有静态分析器和运行时检测器难以捕捉。AllocScope 是面向审计员的静态分析框架，通过建模与分配相关的资金流语义来识别分配操纵漏洞。它构建资金分配图，追踪用户贡献与所得支付之间的关系，并定位导致不公平结果的逻辑。作者在 8000 多个真实合约上评估，取得零假阴性；有经验的审计员用户研究也确认其发现准确、可操作，且易于集成到标准审计流程中。
Flawed funds allocation logic in DeFi contracts can result in disproportionate token distribution, reflecting fundamental errors in how contracts determine and assign user payouts. These issues extend beyond funds allocation logic and affect token distribution mechanisms. Existing defenses such as static analyzers and runtime detectors fail to capture these behaviors because each operation appears valid in isolation. In this work, we present AllocScope, an auditor-centric static analysis framework that identifies allocation manipulation vulnerabilities by modeling allocation-related fund flow semantics and generating vulnerability findings for auditors. AllocScope constructs a funds allocation graph to track the relationship between user contributions and received payouts, and identifies logic that result in unfair outcomes. Evaluated on over 8,000 real-world contracts, AllocScope achieves zero false negatives. A user study with experienced auditors confirms that its findings are accurate, actionable, and easy to integrate into standard auditing workflows.
@inproceedings{PanLZJYS26,
title = {AllocScope: Towards Automated Vetting of Allocation Manipulation in Smart Contracts},
author = {Yu Pan and Jing Li and Qi Zhang and Xifeng Jin and Yuannan Yang and Dan She},
booktitle = {FSE},
year = {2026},
doi = {10.1145/3803437.3805223}
}