<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Workflow on</title><link>https://eunus.dev/categories/workflow/</link><description>Recent content in Workflow on</description><generator>Hugo -- gohugo.io</generator><lastBuildDate>Fri, 05 Jun 2026 00:00:00 +0600</lastBuildDate><atom:link href="https://eunus.dev/categories/workflow/index.xml" rel="self" type="application/rss+xml"/><item><title>Don't Babysit One Coding Agent. Orchestrate a Fleet.</title><link>https://eunus.dev/blog/dont-babysit-one-coding-agent.-orchestrate-a-fleet./</link><pubDate>Fri, 05 Jun 2026 00:00:00 +0600</pubDate><guid>https://eunus.dev/blog/dont-babysit-one-coding-agent.-orchestrate-a-fleet./</guid><description>Have been running parallel coding agents for more than a month now, so here are some of my takes.
It feels less like using a tool and more like changing how I execute the same kinds of tasks. A lot of it is just wiring things together (the skills) and letting agents handle the repetitive parts. There&amp;rsquo;s a natural ceiling here though: beyond 2-3 agents in parallel, it gets harder to keep track of what&amp;rsquo;s actually going on.</description></item></channel></rss>