<?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 Jeanphilo Blog</title><link>https://shio-chan-dev.github.io/jeanblog/zh/tags/agent/</link><description>Recent content in Agent on Jeanphilo Blog</description><generator>Hugo -- 0.165.0</generator><language>zh-cn</language><lastBuildDate>Fri, 28 Aug 2026 00:00:00 +0000</lastBuildDate><atom:link href="https://shio-chan-dev.github.io/jeanblog/zh/tags/agent/index.xml" rel="self" type="application/rss+xml"/><item><title>先做信息对等，再谈模型能力：Agent 失败时该优化系统还是模型</title><link>https://shio-chan-dev.github.io/jeanblog/zh/thoughts/thoughts/information-parity-before-model-optimization/</link><pubDate>Fri, 28 Aug 2026 00:00:00 +0000</pubDate><guid>https://shio-chan-dev.github.io/jeanblog/zh/thoughts/thoughts/information-parity-before-model-optimization/</guid><description>Agent 做错判断时，不要先凭感觉升级模型或增加工作流。先冻结模型实际可见的输入，再用同输入可解性、信息增量和输出唯一性三项测试定位问题究竟在推理侧、系统侧还是任务契约。</description></item></channel></rss>