<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Agentic-Coding on BRYSGO</title><link>https://www.brysgo.com/tags/agentic-coding/</link><description>Recent content in Agentic-Coding on BRYSGO</description><generator>Hugo</generator><language>en-us</language><lastBuildDate>Thu, 30 Jul 2026 14:25:40 +0000</lastBuildDate><atom:link href="https://www.brysgo.com/tags/agentic-coding/index.xml" rel="self" type="application/rss+xml"/><item><title>We Let AI Skip Code Review. Math Didn't.</title><link>https://www.brysgo.com/post/2026-07-30-we-let-ai-skip-code-review-math-didn-t/</link><pubDate>Thu, 30 Jul 2026 14:25:40 +0000</pubDate><guid>https://www.brysgo.com/post/2026-07-30-we-let-ai-skip-code-review-math-didn-t/</guid><description>&lt;h2 id="the-gap-between-it-compiles-and-its-right"&gt;The Gap Between &amp;ldquo;It Compiles&amp;rdquo; and &amp;ldquo;It&amp;rsquo;s Right&amp;rdquo;&lt;/h2&gt;
&lt;p&gt;Every AI coding agent I use now writes, tests, and merges code faster than I can read it, and I&amp;rsquo;ve noticed I stopped actually reading it. I skim the diff, I see green checkmarks, I approve. That&amp;rsquo;s not a confession of laziness — it&amp;rsquo;s what the workflow trains you to do. The agent writes the test, the agent writes the code that passes the test, and I&amp;rsquo;m left rubber-stamping a system that graded its own homework.&lt;/p&gt;</description></item></channel></rss>