arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2605.28071cs.CR
收藏

AgentGuard:面向工具使用型LLM智能体的基于属性的访问控制框架

AgentGuard: An Attribute-Based Access Control Framework for Tool-Use LLM-Based Agent

Fudan University(复旦大学);Shanghai Innovation Institute(上海创新研究院)

Jiaqi Luo, Songyang Peng, Jiarun Dai, Zhile Chen, Zhuoxiang Shen, Geng Hong, Xudong Pan, Yuan Zhang, Min Yang

AI总结:

提出AgentGuard框架,通过客户端-服务器架构和三种互补检查机制,为工具使用型LLM智能体提供轻量级集成与跨工具安全风险防护。

摘要:

基于LLM的智能体因其能够自主调用相关工具完成复杂任务而近期受到广泛关注。然而,研究表明这些智能体面临严重的安全风险,可能导致隐私泄露、财务损失甚至系统完全受损。本文提出AgentGuard,一个面向工具使用型LLM智能体的基于属性的访问控制框架。AgentGuard采用客户端-服务器架构。在客户端,AgentGuard为不同编程语言和架构实现的智能体提供轻量级集成,仅需少量代码修改(例如约10行),且不改变底层智能体执行逻辑。在服务器端,AgentGuard提供三种互补的检查机制,覆盖智能体执行中的单工具和跨工具安全风险。此外,它还提供可视化的前端界面用于安全策略规范和运行时审计。目前,AgentGuard已在https://github.com/WhitzardAgent/AgentGuard公开访问。

英文摘要:

LLM-based agents have recently attracted significant attention due to their ability to autonomously invoke relevant tools to accomplish complex tasks. However, recent studies have shown that these agents face severe security risks, which may lead to privacy leakage, financial loss, or even full system compromise. In this paper, we present AgentGuard, an attribute-based access control framework for tool-use LLM-based agents. AgentGuard adopts a client-server architecture. On the client side, AgentGuard provides lightweight integration for agents implemented in different programming languages and architectures. It requires only minor code modifications (e.g., around 10 lines) without changing the underlying agent execution logic. On the server side, AgentGuard provides three complementary inspection mechanisms to cover both single-tool and cross-tool security risks in agent execution. In addition, it offers a visualized front-end interface for security policy specification and runtime auditing. Currently, AgentGuard is publicly accessible at https://github.com/WhitzardAgent/AgentGuard.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑