<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Jailbreaks AI - Blog</title>
    <link>https://jailbreaksai.com/blog/</link>
    <description>Notes from the lab: how the prompt library is tested, what breaks on a model patch, and the research behind the model tags.</description>
    <language>en-gb</language>
    <lastBuildDate>Fri, 13 Nov 2026 12:46:23 +0000</lastBuildDate>
    <atom:link href="https://jailbreaksai.com/feed.xml" rel="self" type="application/rss+xml"/>
    <item>
      <title>Why prompts break, and how the library tracks it</title>
      <link>https://jailbreaksai.com/blog/why-prompts-break/</link>
      <guid isPermaLink="true">https://jailbreaksai.com/blog/why-prompts-break/</guid>
      <description>A provider ships a patch on a Tuesday and half the library stops working by Wednesday. Here is the process that catches it, and why the model tag is the product.</description>
      <pubDate>Fri, 13 Nov 2026 09:00:00 +0000</pubDate>
    </item>
    <item>
      <title>The thinking-block technique</title>
      <link>https://jailbreaksai.com/blog/the-thinking-block-technique/</link>
      <guid isPermaLink="true">https://jailbreaksai.com/blog/the-thinking-block-technique/</guid>
      <description>Why the strongest long-form prompts in the library all insist on a continuous thinking block, and what happens when you truncate one.</description>
      <pubDate>Fri, 13 Nov 2026 09:00:00 +0000</pubDate>
    </item>
    <item>
      <title>How we verify a model tag</title>
      <link>https://jailbreaksai.com/blog/how-we-verify-a-model-tag/</link>
      <guid isPermaLink="true">https://jailbreaksai.com/blog/how-we-verify-a-model-tag/</guid>
      <description>The October evaluation run: 17 models, 23 tested, scored out of 25 on five axes. What the numbers do and do not tell you.</description>
      <pubDate>Fri, 13 Nov 2026 09:00:00 +0000</pubDate>
    </item>
  </channel>
</rss>
