<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/"><channel><title>Raja Sleem</title><link>https://www.rajasleem.com/</link><description>Recent content on Raja Sleem</description><image><title>Raja Sleem</title><url>https://www.rajasleem.com/images/rajasleem-og-thumbnail-1200x630.png</url><link>https://www.rajasleem.com/images/rajasleem-og-thumbnail-1200x630.png</link></image><generator>Hugo</generator><language>en-us</language><lastBuildDate>Wed, 15 Jul 2026 09:00:00 -0400</lastBuildDate><atom:link href="https://www.rajasleem.com/index.xml" rel="self" type="application/rss+xml"/><item><title>Letting It Run: Toward an Autonomous Build Loop</title><link>https://www.rajasleem.com/posts/agentic-workflow-part-4-letting-it-run/</link><pubDate>Wed, 15 Jul 2026 09:00:00 -0400</pubDate><guid>https://www.rajasleem.com/posts/agentic-workflow-part-4-letting-it-run/</guid><description>Once agents could build to a shared contract and be checked against it, I stopped driving every step by hand. This is the loop that runs a batch of features mostly on its own.</description></item><item><title>When Agents Diverge: The Code Contract</title><link>https://www.rajasleem.com/posts/agentic-workflow-part-3-when-agents-diverge/</link><pubDate>Mon, 06 Jul 2026 09:00:00 -0400</pubDate><guid>https://www.rajasleem.com/posts/agentic-workflow-part-3-when-agents-diverge/</guid><description>Run several agents on the same project and they each build something reasonable that doesn&amp;#39;t match the others. Here&amp;#39;s the fix: write the rules down once, before anyone builds.</description></item><item><title>Surviving the Context Window</title><link>https://www.rajasleem.com/posts/agentic-workflow-part-2-surviving-the-context-window/</link><pubDate>Sun, 07 Jun 2026 09:00:00 -0400</pubDate><guid>https://www.rajasleem.com/posts/agentic-workflow-part-2-surviving-the-context-window/</guid><description>One Claude session is never enough for real work. This is how I hand off between sessions, push work onto sub-agents to save context, and finally see all the planning files at a glance.</description></item><item><title>From Prompts to a Process: Making Claude Code Repeatable</title><link>https://www.rajasleem.com/posts/agentic-workflow-part-1-from-prompts-to-a-process/</link><pubDate>Mon, 01 Jun 2026 09:00:00 -0400</pubDate><guid>https://www.rajasleem.com/posts/agentic-workflow-part-1-from-prompts-to-a-process/</guid><description>The first thing I fixed about working with an LLM wasn&amp;#39;t the model. It was me starting from scratch every session. Here&amp;#39;s the simple process that changed that.</description></item><item><title>AI Fatigue Is Real</title><link>https://www.rajasleem.com/posts/ai-fatigue/</link><pubDate>Sun, 31 May 2026 12:00:00 -0400</pubDate><guid>https://www.rajasleem.com/posts/ai-fatigue/</guid><description>The tools got faster. My brain didn&amp;#39;t. Why four hours of building with an LLM wipes me out, and why I doubt the &amp;#39;spin up 200 agents&amp;#39; future.</description></item></channel></rss>