<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Defense in Depth on Vivian＠資安媽媽包</title><link>https://vvnblog.com/tags/defense-in-depth/</link><description>Recent content in Defense in Depth on Vivian＠資安媽媽包</description><generator>Hugo -- gohugo.io</generator><language>zh-tw</language><copyright>Copyright © 2026 Vivian All Rights Reserved.</copyright><lastBuildDate>Mon, 14 Sep 2026 00:00:00 +0800</lastBuildDate><atom:link href="https://vvnblog.com/tags/defense-in-depth/index.xml" rel="self" type="application/rss+xml"/><item><title>AI 黑魔法（26）：縱深防禦（二），Agent、MCP 跟模型檔的防線</title><link>https://vvnblog.com/posts/ai-dark-arts-26/</link><pubDate>Mon, 14 Sep 2026 00:00:00 +0800</pubDate><guid>https://vvnblog.com/posts/ai-dark-arts-26/</guid><description>看 OWASP 第 8 到 10 條要你在哪裡限制權限、記憶誰能寫、外面接進來的東西要驗什麼。</description></item><item><title>AI 黑魔法（25）：縱深防禦（一），Prompt Injection 防護要做在哪幾層？</title><link>https://vvnblog.com/posts/ai-dark-arts-25/</link><pubDate>Sun, 13 Sep 2026 00:00:00 +0800</pubDate><guid>https://vvnblog.com/posts/ai-dark-arts-25/</guid><description>根據 OWASP LLM01:2026 的 11 條建議，看 Prompt Injection 防護每一層擋得住什麼，漏過去之後又靠什麼。</description></item></channel></rss>