<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/"><channel><title>Ai-Agents on socratic notes</title><link>https://socraticagent.dev/tags/ai-agents/</link><description>Recent content in Ai-Agents on socratic notes</description><generator>Hugo</generator><language>en</language><lastBuildDate>Wed, 12 Aug 2026 00:00:00 +0000</lastBuildDate><atom:link href="https://socraticagent.dev/tags/ai-agents/index.xml" rel="self" type="application/rss+xml"/><item><title>The Guardrail Has to Live Outside the Model</title><link>https://socraticagent.dev/blog/the-guardrail-has-to-live-outside-the-model/</link><pubDate>Wed, 12 Aug 2026 00:00:00 +0000</pubDate><guid>https://socraticagent.dev/blog/the-guardrail-has-to-live-outside-the-model/</guid><description>Why a safety check for an AI agent can never live inside the same model call it&amp;#39;s supposed to govern, and what a real four-checkpoint guardrail pipeline looks like in practice.</description></item></channel></rss>