What should a developer inspect before deploying an LLM agent: the model, the tool code, the deployment configuration, or all three? In practice, many security failures in agent systems arise not from model weights alone, but from the surrounding software stack: tool functions that pass untrusted inputs to dangerous operations, exposed credentials in deployment artifacts, and over-privileged Model Context Protocol (MCP) configurations. We present Agent Audit, a security analysis system for LLM agent applications. Agent Audit analyzes Python agent code and deployment artifacts through an agent-aware pipeline that combines dataflow analysis, credential detection, structured configuration parsing, and privilege-risk checks. The system reports findings in terminal, JSON, and SARIF formats, enabling direct integration with local development workflows and CI/CD pipelines. On a benchmark of 22 samples with 42 annotated vulnerabilities, Agent Audit detects 40 vulnerabilities with 6 false positives, substantially improving recall over common SAST baselines while maintaining sub-second scan times. Agent Audit is open source and installable via pip, making security auditing accessible for agent systems. In the live demonstration, attendees scan vulnerable agent repositories and observe how Agent Audit identifies security risks in tool functions, prompts, and more. Findings are linked to source locations and configuration paths, and can be exported into VS Code and GitHub Code Scanning for interactive inspection.

翻译：开发者在部署大语言模型智能体之前应检查哪些内容：模型、工具代码、部署配置，还是三者全部？实践中，智能体系统的许多安全故障不仅源于模型权重本身，更来自其周围的软件栈：将不可信输入传递给危险操作的工具函数、部署产物中暴露的凭证信息，以及权限过高的模型上下文协议配置。我们提出Agent Audit，一种面向LLM智能体应用的安全分析系统。Agent Audit通过融合数据流分析、凭证检测、结构化配置解析和权限风险检查的智能体感知流水线，对Python智能体代码及部署产物进行分析。该系统支持终端、JSON和SARIF格式的结果输出，可直接集成至本地开发工作流与CI/CD流水线。在包含22个样本、42个标注漏洞的基准测试中，Agent Audit检测出40个漏洞（含6个误报），在保持亚秒级扫描速度的同时，显著提升了相较于常规SAST基线的召回率。Agent Audit以开源形式发布，可通过pip安装，使智能体系统的安全审计触手可及。在实时演示环节，参与者可扫描存在漏洞的智能体仓库，观察Agent Audit如何识别工具函数、提示词等环节的安全风险。检测结果关联至源代码位置与配置路径，并可导出至VS Code和GitHub Code Scanning进行交互式审查。