<?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>Agent on shinu</title><link>https://shinu.dev/blog/tags/agent/</link><description>Recent content in Agent on shinu</description><generator>Hugo</generator><language>zh-cn</language><lastBuildDate>Thu, 29 Oct 2026 22:02:23 +0800</lastBuildDate><atom:link href="https://shinu.dev/blog/tags/agent/index.xml" rel="self" type="application/rss+xml"/><item><title>理解 LLM Agent Loop：智能体循环的七个步骤</title><link>https://shinu.dev/blog/posts/llm-agent-loop/</link><pubDate>Thu, 29 Oct 2026 22:02:23 +0800</pubDate><guid>https://shinu.dev/blog/posts/llm-agent-loop/</guid><description>&lt;p&gt;LLM Agent 的核心运行方式，可以理解为一个不断循环的过程：感知信息、推理判断、规划行动、执行操作、观察结果、反思调整，直到任务完成或终止。&lt;/p&gt;
&lt;p&gt;下面整理 Agent Loop 的七个步骤。&lt;/p&gt;</description></item></channel></rss>