<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>AgentiLoop Agent! Blog</title>
    <link>https://agentiloop.ai/blog/</link>
    <atom:link href="https://agentiloop.ai/blog/feed.xml" rel="self" type="application/rss+xml"/>
    <description>Deep dives into AgentiLoop Agent!, the native macOS AI agent.</description>
    <language>en</language>
    <lastBuildDate>Wed, 30 Sep 2026 12:00:00 +0000</lastBuildDate>
    <image>
      <url>https://agentiloop.ai/agent_icon.png</url>
      <title>AgentiLoop Agent! Blog</title>
      <link>https://agentiloop.ai/blog/</link>
    </image>
    <item>
      <title>Why Agent! Compacts at Half the Window, and the Three Bugs That Shrank It to 16K</title>
      <link>https://agentiloop.ai/blog/context-compaction-half-the-window/</link>
      <guid isPermaLink="true">https://agentiloop.ai/blog/context-compaction-half-the-window/</guid>
      <pubDate>Wed, 30 Sep 2026 12:00:00 +0000</pubDate>
      <dc:creator>Todd Bruss</dc:creator>
      <category>Internals</category>
      <category>Release Notes</category>
      <description>How Agent!&#x27;s context compaction decides when to summarize a long task, and the September 28 fixes for fallback models, Ollama and a 272K Codex window.</description>
    </item>
    <item>
      <title>A Release on the First of Every Month, and Pre-Releases in Between</title>
      <link>https://agentiloop.ai/blog/a-release-on-the-first-of-every-month/</link>
      <guid isPermaLink="true">https://agentiloop.ai/blog/a-release-on-the-first-of-every-month/</guid>
      <pubDate>Tue, 29 Sep 2026 12:00:00 +0000</pubDate>
      <dc:creator>Todd Bruss</dc:creator>
      <category>Release Notes</category>
      <category>Behind the Scenes</category>
      <description>Agent! now ships a stable release on the 1st of every month. Daily pre-releases (moving to weekly) are where new features and fixes get tried out, and release candidates lock things down before the big day.</description>
    </item>
    <item>
      <title>The Terminal Strikes Back: Taking the Agent Loop Cross-Platform in Rust and Go</title>
      <link>https://agentiloop.ai/blog/the-terminal-strikes-back/</link>
      <guid isPermaLink="true">https://agentiloop.ai/blog/the-terminal-strikes-back/</guid>
      <pubDate>Mon, 28 Sep 2026 12:00:00 +0000</pubDate>
      <dc:creator>Todd Bruss</dc:creator>
      <category>Announcement</category>
      <category>Cross-Platform</category>
      <category>Engineering</category>
      <description>Agent! is a Mac app, but a coding agent belongs wherever developers work. Here&#x27;s how the agent loop became two CLIs, one in Rust and one in Go, with a full-screen TUI on macOS, Windows and Linux.</description>
    </item>
    <item>
      <title>Agent! Now Runs on macOS 14.6 and Intel Macs: How We Got Off macOS 26</title>
      <link>https://agentiloop.ai/blog/agent-now-runs-on-macos-14-6-and-intel/</link>
      <guid isPermaLink="true">https://agentiloop.ai/blog/agent-now-runs-on-macos-14-6-and-intel/</guid>
      <pubDate>Sun, 27 Sep 2026 12:00:00 +0000</pubDate>
      <dc:creator>Todd Bruss</dc:creator>
      <category>Release Notes</category>
      <category>Engineering</category>
      <description>Agent! was built around Apple Intelligence on macOS 26. Here&#x27;s how one day of #available gates and package bumps brought it to macOS 14.6 on Apple Silicon and Intel.</description>
    </item>
    <item>
      <title>Before the Model Gets a Vote: Inside Agent!&#x27;s Shell Guardrails</title>
      <link>https://agentiloop.ai/blog/inside-agent-shell-guardrails/</link>
      <guid isPermaLink="true">https://agentiloop.ai/blog/inside-agent-shell-guardrails/</guid>
      <pubDate>Sat, 26 Sep 2026 12:00:00 +0000</pubDate>
      <dc:creator>Todd Bruss</dc:creator>
      <category>Security</category>
      <category>Internals</category>
      <description>How ShellSafetyService, the daemon-side re-check and the Jev second opinion stop an AI agent from wiping your Mac, walked through the actual Swift code.</description>
    </item>
    <item>
      <title>Welcome to the Agent! Blog: One App, Any AI, Total Command Over Your Mac</title>
      <link>https://agentiloop.ai/blog/welcome-to-the-agent-blog/</link>
      <guid isPermaLink="true">https://agentiloop.ai/blog/welcome-to-the-agent-blog/</guid>
      <pubDate>Fri, 25 Sep 2026 12:00:00 +0000</pubDate>
      <dc:creator>Todd Bruss</dc:creator>
      <category>Announcement</category>
      <category>Architecture</category>
      <description>What AgentiLoop Agent! is, how it&#x27;s built, and what this daily blog will cover, straight from the source code.</description>
    </item>
    <item>
      <title>Meet Jev: A Second Opinion That Answers in Types, Not Prose</title>
      <link>https://agentiloop.ai/blog/meet-jev/</link>
      <guid isPermaLink="true">https://agentiloop.ai/blog/meet-jev/</guid>
      <pubDate>Thu, 17 Sep 2026 12:00:00 +0000</pubDate>
      <dc:creator>Todd Bruss</dc:creator>
      <category>Security</category>
      <category>Announcement</category>
      <description>Agent! now asks Jev (TypeSafe System One) how destructive a shell command is before running it. Here&#x27;s why Jev is an advisor and not another LLM provider.</description>
    </item>
    <item>
      <title>Why We Ripped Out Coding Mode: Let the Model Pick Its Own Tools</title>
      <link>https://agentiloop.ai/blog/why-we-ripped-out-modes/</link>
      <guid isPermaLink="true">https://agentiloop.ai/blog/why-we-ripped-out-modes/</guid>
      <pubDate>Tue, 07 Apr 2026 12:00:00 +0000</pubDate>
      <dc:creator>Todd Bruss</dc:creator>
      <category>Architecture</category>
      <category>Design</category>
      <description>Agent! used to guess whether you wanted coding or automation and hide tools to match. One broken Photo Booth task showed why the harness should stop second-guessing the model.</description>
    </item>
  </channel>
</rss>
