**We're securing Tabstack against indirect prompt injection**

> A technical look at the prompt injection vulnerability in Tabstack and the code changes made to fix it. At Mozilla, we believe that building a useful AI ecosystem requires radical transparency, especially when it comes to security. Recently, security researchers at Brave reached out to us regarding an Indirect Prompt Injection (IPI) vulnerability they identified in Tabstack&\#x27;s /v1/automate endpoint, which they ha...

**来源信息**
- **来源**：Hacker News：AI 热帖
- **分类**：行业
- **发布时间**：2026-06-22 22:18（北京时间）
- **原文**：[打开原文](https://tabstack.ai/blog/securing-ai-agents-indirect-prompt-injection)