<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Architecture on Yesterday I was wrong</title>
    <link>/tags/architecture/</link>
    <description>Recent content in Architecture on Yesterday I was wrong</description>
    <generator>Hugo</generator>
    <language>en-gb</language>
    <lastBuildDate>Wed, 29 Apr 2026 09:00:00 +0000</lastBuildDate>
    <atom:link href="/tags/architecture/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>Agentic Architecture</title>
      <link>/2026/04/29/agentic-architecture/</link>
      <pubDate>Wed, 29 Apr 2026 09:00:00 +0000</pubDate>
      <guid>/2026/04/29/agentic-architecture/</guid>
      <description>&lt;p&gt;This is part 4 of a 5-part series on agentic engineering. &lt;a href=&#34;/2026/04/26/from-agile-engineering-to-agentic-engineering/&#34;&gt;Part 1&lt;/a&gt; set the values-vs-mechanisms frame. &lt;a href=&#34;/2026/04/27/from-pair-programming-to-co-steering/&#34;&gt;Part 2&lt;/a&gt; covered pairing and learning. &lt;a href=&#34;/2026/04/28/harness-engineering-building-a-factory-for-code/&#34;&gt;Part 3&lt;/a&gt; made the case for harness engineering. This one is about the architecture &lt;em&gt;underneath&lt;/em&gt; the harness — the language, runtime, deployment, and version-control choices that make agents safe.&lt;/p&gt;&#xA;&lt;p&gt;&lt;a href=&#34;/2026/04/30/picking-a-language-for-agents/&#34;&gt;Part 5&lt;/a&gt; gets concrete about specific languages.&lt;/p&gt;&#xA;&lt;h2 id=&#34;why-architecture-changes-with-agents&#34;&gt;Why architecture changes with agents&lt;/h2&gt;&#xA;&lt;p&gt;If humans are no longer the primary readers and writers of code, then architecture should no longer be optimised only for human familiarity. Instead, I want to optimise for:&lt;/p&gt;</description>
    </item>
  </channel>
</rss>
