<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>mko.run</title>
    <link>https://mko.run/</link>
    <atom:link href="https://mko.run/feed.xml" rel="self" type="application/rss+xml"/>
    <language>en</language>
    <item>
      <title>Firecracker on WSL2, without root</title>
      <link>https://mko.run/firecracker-wsl2-rootless/</link>
      <guid isPermaLink="true">https://mko.run/firecracker-wsl2-rootless/</guid>
      <pubDate>Tue, 10 Nov 2026 00:00:00 GMT</pubDate>
      <description>How I run each coding agent in its own Firecracker microVM nested in WSL2, with no root at run time: the stack, what it costs, and ten stumbling blocks.</description>
    </item>
    <item>
      <title>Credential proxies fail open in the details</title>
      <link>https://mko.run/default-deny/</link>
      <guid isPermaLink="true">https://mko.run/default-deny/</guid>
      <pubDate>Tue, 03 Nov 2026 00:00:00 GMT</pubDate>
      <description>Where credential proxies for AI coding agents fail open: Docker's and microsandbox's advisories of autumn 2026, three decisions a proxy makes, and what d0 does at each.</description>
    </item>
    <item>
      <title>EmbeddingGemma in Go, tokenizer and all</title>
      <link>https://mko.run/embeddinggemma-in-go/</link>
      <guid isPermaLink="true">https://mko.run/embeddinggemma-in-go/</guid>
      <pubDate>Mon, 02 Nov 2026 00:00:00 GMT</pubDate>
      <description>How “ask teblo” matches questions with EmbeddingGemma in plain Go: the models measured, the tokenizer, the one output that matters.</description>
    </item>
  </channel>
</rss>
