https://reason.pgch.app/Recent content on Reason and ActHugoenThu, 29 Oct 2026 02:30:00 +0700https://reason.pgch.app/posts/build-your-own-agent-harness/Thu, 29 Oct 2026 02:30:00 +0700https://reason.pgch.app/posts/build-your-own-agent-harness/<p>The first project I shipped after Claude Opus 4.5 came out was my own agent harness.</p> <p>Opus 4.5 came out on 2025-11-24. For many developers, me included, it was the point where coding agents started to work for real. Ethan Mollick had early access and posted this on launch day:</p> <div class="xpost-embed"> <blockquote class="twitter-tweet" data-dnt="true" data-conversation="none"><p lang="en" dir="ltr">I had early access to Opus 4.5 &amp; it is a very impressive model that seem to be right at the frontier Big gains in ability to do practical work (like make a PowerPoint from an Excel) and the best results ever (&amp; in one shot) in my Lem poetry test, plus good results in Claude Code</p>&mdash; Ethan Mollick (@emollick) <a href="https://x.com/emollick/status/1993030988759470156">November 24, 2025</a></blockquote> </div>https://reason.pgch.app/about/Mon, 01 Jan 0001 00:00:00 +0000https://reason.pgch.app/about/<p>Like most developers who got one-shotted by Claude Opus 4.5 last December, I have noticed a steady decline in my ability to write structured, coherent thoughts over the past few months. This blog is my attempt to rebuild the habit of long-form, structured thinking.</p> <p>I will start with the topics I care about:</p> <ul> <li><strong>Agentic engineering</strong>: building, running and measuring agent harnesses.</li> <li><strong>Homelab</strong>: the machines, network and services that the agents run on.</li> <li><strong>Local LLMs</strong>: open-weights models on my own hardware, and what they can actually do.</li> </ul>