<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Software Development on Pfisterer Consulting</title><link>https://pfisterer.xyz/en/tags/software-development/</link><description>Recent content in Software Development on Pfisterer Consulting</description><generator>Hugo</generator><language>en</language><lastBuildDate>Tue, 31 Mar 2026 14:30:00 +0200</lastBuildDate><atom:link href="https://pfisterer.xyz/en/tags/software-development/index.xml" rel="self" type="application/rss+xml"/><item><title>Context Rot in AI Agents: Why Quality Degrades</title><link>https://pfisterer.xyz/en/news/gsd-get-shit-done-ki-entwicklung-context-engineering/</link><pubDate>Tue, 31 Mar 2026 14:30:00 +0200</pubDate><guid>https://pfisterer.xyz/en/news/gsd-get-shit-done-ki-entwicklung-context-engineering/</guid><description>&lt;p&gt;Anyone who works with Claude Code, Cursor, or Copilot knows the pattern. The problem has a name: &lt;strong&gt;Context Rot&lt;/strong&gt;. The first results are sharp and the architecture holds up. Then quality drops off. After 20, 30 minutes, the answers start repeating themselves. The AI agent loses context. It produces code that contradicts what it wrote five prompts ago.&lt;/p&gt;
&lt;p&gt;&lt;strong&gt;Not a user error &amp;ndash; an architecture problem.&lt;/strong&gt; The solution lies in context engineering: the deliberate management of context windows through multi-agent orchestration. These are exactly the kinds of real-world problems I address in my &lt;a href="https://pfisterer.xyz/en/leistungen/ki-automatisierung/"&gt;AI &amp;amp; Automation consulting&lt;/a&gt;.&lt;/p&gt;</description></item></channel></rss>