<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Selfhosted on Underkube</title><link>https://www.underkube.com/tags/selfhosted/</link><description>Recent content in Selfhosted on Underkube</description><generator>Hugo</generator><language>en-us</language><lastBuildDate>Sat, 20 Jun 2026 07:00:00 +0200</lastBuildDate><atom:link href="https://www.underkube.com/tags/selfhosted/index.xml" rel="self" type="application/rss+xml"/><item><title>What Edu is reading this week (Jun 14 - 20, 2026)</title><link>https://www.underkube.com/2026-06-20-what-edu-is-reading-this-week-jun-14-20-2026/</link><pubDate>Sat, 20 Jun 2026 07:00:00 +0200</pubDate><guid>https://www.underkube.com/2026-06-20-what-edu-is-reading-this-week-jun-14-20-2026/</guid><description>A week heavy on local LLM coding setups, token-trimming tooling, dotfile managers, ZFS self-hosting, and a couple of supply-chain security reminders.
Local AI &amp;amp; Coding Agents Running local models is good now / Ask HN: Has anyone replaced Claude/GPT with a local model?: Vicki Boykis argues local agentic coding has become genuinely good over the past few months, alongside the Hacker News thread of people sharing their daily-driver local setups. How to Setup a Local Coding Agent on macOS: Running Gemma 4 26B-A4B and Qwen3.</description></item></channel></rss>