<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>_Atomizations on John Young</title><link>https://jyoung.dev/_atomizations/</link><description>Recent content in _Atomizations on John Young</description><generator>Hugo</generator><language>en-us</language><lastBuildDate>Mon, 11 May 2026 00:00:00 +0000</lastBuildDate><atom:link href="https://jyoung.dev/_atomizations/index.xml" rel="self" type="application/rss+xml"/><item><title/><link>https://jyoung.dev/_atomizations/how-to-size-tasks-for-ai-coding-agents/</link><pubDate>Mon, 11 May 2026 00:00:00 +0000</pubDate><guid>https://jyoung.dev/_atomizations/how-to-size-tasks-for-ai-coding-agents/</guid><description>&lt;h2 id="1-main-thesis-250-words">1. Main thesis (~250 words)&lt;/h2>
&lt;p>The real constraint on AI agent task size isn&amp;rsquo;t lines of code. It&amp;rsquo;s context.&lt;/p>
&lt;p>Most people size tasks by gut feel — &amp;ldquo;that seems about right&amp;rdquo; — and then wonder why the agent spirals into corrections halfway through. Lines changed and files touched are secondary proxies. The actual limiter is how much context the agent burns reading, exploring, and running commands before it can start writing the diff.&lt;/p></description></item></channel></rss>