<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Ai-Coding on Kostiantyn Lysenko</title><link>https://lysenko.dev/tags/ai-coding/</link><description>Recent content in Ai-Coding on Kostiantyn Lysenko</description><generator>Hugo -- gohugo.io</generator><language>en</language><copyright>© 2026 Kostiantyn Lysenko</copyright><lastBuildDate>Sun, 10 May 2026 15:00:00 +0700</lastBuildDate><atom:link href="https://lysenko.dev/tags/ai-coding/index.xml" rel="self" type="application/rss+xml"/><item><title>Trimming the chatter from Claude Code with the caveman plugin</title><link>https://lysenko.dev/posts/2026-05-caveman-plugin-claude-code/</link><pubDate>Sun, 10 May 2026 15:00:00 +0700</pubDate><guid>https://lysenko.dev/posts/2026-05-caveman-plugin-claude-code/</guid><description>&lt;p&gt;Most LLM answers come wrapped in throat-clearing. &amp;ldquo;Sure! I&amp;rsquo;d be happy to help. The reason your component re-renders is most likely&amp;hellip;&amp;rdquo; By the time the useful sentence arrives, I&amp;rsquo;ve read four lines of nothing.&lt;/p&gt;
&lt;p&gt;The cost is mental load, not just dollars. Every reply forces me to skim past pleasantries for the answer. Across a session, the skimming compounds into fatigue.&lt;/p&gt;</description></item></channel></rss>