<?xml version="1.0" encoding="UTF-8"?>
<feed xmlns="http://www.w3.org/2005/Atom" xml:lang="en">
    <title>AgenticLinux - Blog</title>
    <subtitle>A Linux desktop built for working with AI agents. The claude, codex, opencode, openclaw and goose agents, Docker, Docker Sandboxes, QEMU with KVM and GPU runtimes are there on first boot, in a bootc image with atomic updates and automatic rollback.</subtitle>
    <link rel="self" type="application/atom+xml" href="https://agenticlinux.github.io/blog/atom.xml"/>
    <link rel="alternate" type="text/html" href="https://agenticlinux.github.io/blog/"/>
    <generator uri="https://www.getzola.org/">Zola</generator>
    <updated>2026-11-04T00:00:00+00:00</updated>
    <id>https://agenticlinux.github.io/blog/atom.xml</id>
    <entry xml:lang="en">
        <title>AgenticLinux on Apple Silicon: the 16k variants</title>
        <published>2026-11-04T00:00:00+00:00</published>
        <updated>2026-11-04T00:00:00+00:00</updated>
        
        <author>
          <name>Unknown</name>
        </author>
        
        <link rel="alternate" type="text/html" href="https://agenticlinux.github.io/blog/apple-silicon-16k-variants/"/>
        <id>https://agenticlinux.github.io/blog/apple-silicon-16k-variants/</id>
        
        <summary type="html">&lt;p&gt;AgenticLinux now has variants for Apple Silicon Macs. On a Mac mini M1 that
we switched to the KDE one, the kernel reports a 16 KiB page size, the
Plasma login screen comes up, the Wi-Fi joins on its own, the GPU shows up
in Vulkan as &lt;code&gt;Apple M1 (G13G B1)&lt;/code&gt; on the Honeykrisp driver, and
&lt;code&gt;claude&lt;/code&gt;, &lt;code&gt;codex&lt;/code&gt;, &lt;code&gt;openclaw&lt;/code&gt; and &lt;code&gt;goose&lt;/code&gt; start.&lt;/p&gt;
&lt;p&gt;This post covers what the &lt;code&gt;16k-&lt;/code&gt; variants are, how they are built, how to
switch to one, what went wrong on the way, and what we did not test.&lt;/p&gt;</summary>
        
    </entry>
    <entry xml:lang="en">
        <title>AgenticLinux on the NVIDIA DGX Spark</title>
        <published>2026-11-04T00:00:00+00:00</published>
        <updated>2026-11-04T00:00:00+00:00</updated>
        
        <author>
          <name>Unknown</name>
        </author>
        
        <link rel="alternate" type="text/html" href="https://agenticlinux.github.io/blog/dgx-spark-support/"/>
        <id>https://agenticlinux.github.io/blog/dgx-spark-support/</id>
        
        <summary type="html">&lt;p&gt;AgenticLinux now runs on the NVIDIA DGX Spark. On a Spark running the KDE
variant, &lt;code&gt;nvidia-smi&lt;/code&gt; sees the GB10 GPU with the 580.196.07 driver, the
Wi-Fi, NetworkManager and &lt;code&gt;sshd&lt;/code&gt; come up on their own, and the agents run on
a local Qwen 3.8 27B.&lt;/p&gt;
&lt;p&gt;This post covers the driver, the agents running on the Spark, how we
installed AgenticLinux over the Spark’s stock Ubuntu, and what we did not
test.&lt;/p&gt;</summary>
        
    </entry>
    <entry xml:lang="en">
        <title>Introducing AgenticLinux</title>
        <published>2026-11-04T00:00:00+00:00</published>
        <updated>2026-11-04T00:00:00+00:00</updated>
        
        <author>
          <name>Unknown</name>
        </author>
        
        <link rel="alternate" type="text/html" href="https://agenticlinux.github.io/blog/introducing-agenticlinux/"/>
        <id>https://agenticlinux.github.io/blog/introducing-agenticlinux/</id>
        
        <summary type="html">&lt;p&gt;AgenticLinux is a Linux desktop built for working with AI agents.
Everything is there on first boot: the &lt;code&gt;claude&lt;/code&gt;, &lt;code&gt;codex&lt;/code&gt;, &lt;code&gt;opencode&lt;/code&gt;,
&lt;code&gt;openclaw&lt;/code&gt; and &lt;code&gt;goose&lt;/code&gt; agents, a way to run them side by side, a way to
watch their usage limits, and the container and GPU tooling they usually
need.&lt;/p&gt;
&lt;p&gt;It is a &lt;a rel=&quot;external&quot; href=&quot;https://bootc-dev.github.io/bootc/&quot;&gt;bootc&lt;/a&gt; image. The whole OS
ships as a container, so updates are atomic and rollback is one command.&lt;/p&gt;</summary>
        
    </entry>
    <entry xml:lang="en">
        <title>Updates that roll themselves back: bootguard</title>
        <published>2026-11-04T00:00:00+00:00</published>
        <updated>2026-11-04T00:00:00+00:00</updated>
        
        <author>
          <name>Unknown</name>
        </author>
        
        <link rel="alternate" type="text/html" href="https://agenticlinux.github.io/blog/updates-that-roll-themselves-back-bootguard/"/>
        <id>https://agenticlinux.github.io/blog/updates-that-roll-themselves-back-bootguard/</id>
        
        <summary type="html">&lt;p&gt;&lt;code&gt;bootc rollback&lt;/code&gt; gets you back to the previous image, but only if you can
still run it. An update that never reaches userspace, or reaches it and
hangs, leaves you at a firmware prompt or an emergency shell instead.&lt;/p&gt;
&lt;p&gt;bootguard is the part of AgenticLinux that handles this. A newly applied
deployment gets a limited number of boot attempts. If it does not come up
healthy within them, the machine returns to the previous deployment by
itself and remembers not to try that image again.&lt;/p&gt;</summary>
        
    </entry>
    <entry xml:lang="en">
        <title>Why /tmp is not a tmpfs</title>
        <published>2026-11-04T00:00:00+00:00</published>
        <updated>2026-11-04T00:00:00+00:00</updated>
        
        <author>
          <name>Unknown</name>
        </author>
        
        <link rel="alternate" type="text/html" href="https://agenticlinux.github.io/blog/why-tmp-is-not-a-tmpfs/"/>
        <id>https://agenticlinux.github.io/blog/why-tmp-is-not-a-tmpfs/</id>
        
        <summary type="html">&lt;p&gt;The Fedora and CentOS Stream images AgenticLinux is built from use a
tmpfs for &lt;code&gt;/tmp&lt;/code&gt;, and AgenticLinux replaces it. &lt;code&gt;/tmp&lt;/code&gt; is &lt;code&gt;/var/tmp&lt;/code&gt;
mounted over &lt;code&gt;/tmp&lt;/code&gt;, so the two are one directory on the root filesystem,
and RAM does not limit it.&lt;/p&gt;
&lt;p&gt;The reason is what happens when a build leaves gigabytes behind. This post
covers the problem, the two small systemd pieces that fix it, and what
you give up.&lt;/p&gt;</summary>
        
    </entry>
</feed>
