<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Posts on Reason and Act</title><link>https://reason.pgch.app/posts/</link><description>Recent content in Posts on Reason and Act</description><generator>Hugo</generator><language>en</language><lastBuildDate>Thu, 29 Oct 2026 02:30:00 +0700</lastBuildDate><atom:link href="https://reason.pgch.app/posts/index.xml" rel="self" type="application/rss+xml"/><item><title>Everyone is making an agent harness, you should build another one (for yourself)</title><link>https://reason.pgch.app/posts/build-your-own-agent-harness/</link><pubDate>Thu, 29 Oct 2026 02:30:00 +0700</pubDate><guid>https://reason.pgch.app/posts/build-your-own-agent-harness/</guid><description>&lt;p&gt;The first project I shipped after Claude Opus 4.5 came out was my own agent&#10;harness.&lt;/p&gt;&#10;&lt;p&gt;Opus 4.5 came out on 2025-11-24. For many developers, me included, it was the&#10;point where coding agents started to work for real. Ethan Mollick had early&#10;access and posted this on launch day:&lt;/p&gt;&#10;&#10;&lt;div class="xpost-embed"&gt;&#10;&lt;blockquote class="twitter-tweet" data-dnt="true" data-conversation="none"&gt;&lt;p lang="en" dir="ltr"&gt;I had early access to Opus 4.5 &amp;amp; it is a very impressive model that seem to be right at the frontier&#10;&#10;Big gains in ability to do practical work (like make a PowerPoint from an Excel) and the best results ever (&amp;amp; in one shot) in my Lem poetry test, plus good results in Claude Code&lt;/p&gt;&amp;mdash; Ethan Mollick (@emollick) &lt;a href="https://x.com/emollick/status/1993030988759470156"&gt;November 24, 2025&lt;/a&gt;&lt;/blockquote&gt;&#10;&lt;/div&gt;</description></item></channel></rss>