exfil-scan

exfil-scan 扫描LLM输出和AI生成内容中的数据传输信号，在到达用户、日志或下游系统之前进行检测。 1 为什么需要exfil-scan 现代LLM应用通常处理敏感提示、内部文档、API响应和用户数据。 这创建了一个新的安全边界：输出层 。 EchoLeak系列攻击（包括CVE-2025-32711）展示了隐藏或混淆的有效载荷如何被 嵌入在生成的文本中，并通过工具、链接或解析链来窃取数据。 传统扫描器关注文件、依赖项和运行时流量。 它们不会以专门构建的方式检查生成的模型输出。 exfil-scan提供了一个“LLM响应的病毒扫描器”： 检查纯文本、JSON、Markdown和日志...

CVE-2026-72742

DSPy 3.3.0b1 contains a file exfiltration vulnerability in the Image and Audio output field adapters that allows attackers with influence over language model outputs to read arbitrary local files by injecting a filesystem path into the url field of a parsed Image or Audio typed output. The...

CVE-2024-57783

The desktop application in Dot through 0.9.3 allows XSS and resultant command execution because user input and LLM output are appended to the DOM with innerHTML in render.js, and because the Electron window can access Node.js APIs...

GHSA-2HMP-5WQG-F24H PlotAI eval vulnerability

A vulnerability, that could result in Remote Code Execution RCE, has been found in PlotAI. Lack of validation of LLM-generated output allows attacker to execute arbitrary Python code. PlotAI commented out vulnerable line, further usage of the software requires uncommenting it and thus accepting t...

CVE-2024-21552

All versions of SuperAGI are vulnerable to Arbitrary Code Execution due to unsafe use of the ‘eval’ function. An attacker could induce the LLM output to exploit this vulnerability and gain arbitrary code execution on the SuperAGI application server...

CVE-2024-21552

All versions of SuperAGI are vulnerable to Arbitrary Code Execution due to unsafe use of the ‘eval’ function. An attacker could induce the LLM output to exploit this vulnerability and gain arbitrary code execution on the SuperAGI application server...

CVE-2024-21552

All versions of SuperAGI are vulnerable to Arbitrary Code Execution due to unsafe use of the ‘eval’ function. An attacker could induce the LLM output to exploit this vulnerability and gain arbitrary code execution on the SuperAGI application server...