<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>The OmniRush blog</title>
    <link>https://omnirush.ai/blog</link>
    <atom:link href="https://omnirush.ai/blog/rss.xml" rel="self" type="application/rss+xml" />
    <description>Walkthroughs, guides and field notes on AI coding agents from the OmniRush team.</description>
    <language>en</language>
    <lastBuildDate>Mon, 19 Oct 2026 08:00:00 GMT</lastBuildDate>
    <item>
      <title>We gave an AI agent a stress test that hangs. Here is how it found the deadlock.</title>
      <link>https://omnirush.ai/blog/ai-agent-fixes-deadlock</link>
      <guid isPermaLink="true">https://omnirush.ai/blog/ai-agent-fixes-deadlock</guid>
      <pubDate>Mon, 19 Oct 2026 08:00:00 GMT</pubDate>
      <description>A walkthrough: an AI agent diagnoses a hanging stress test in a demo repo, explains the lock-order deadlock, then a three-agent swarm fixes and tests it.</description>
      <category>Walkthrough</category>
      <category>Debugging</category>
      <category>Swarm</category>
    </item>
    <item>
      <title>Desktop app or terminal: where should your AI coding agent live?</title>
      <link>https://omnirush.ai/blog/desktop-vs-terminal-ai-coding-agent</link>
      <guid isPermaLink="true">https://omnirush.ai/blog/desktop-vs-terminal-ai-coding-agent</guid>
      <pubDate>Mon, 19 Oct 2026 08:00:00 GMT</pubDate>
      <description>When an AI coding agent works best in a desktop app, and when in the terminal: workspaces and swarms vs pipes, scripts and SSH. With OmniRush as the example.</description>
      <category>Guides</category>
      <category>Desktop app</category>
      <category>CLI</category>
    </item>
    <item>
      <title>Why run your AI coding agent in a sandbox</title>
      <link>https://omnirush.ai/blog/sandboxed-ai-coding-agent</link>
      <guid isPermaLink="true">https://omnirush.ai/blog/sandboxed-ai-coding-agent</guid>
      <pubDate>Mon, 19 Oct 2026 08:00:00 GMT</pubDate>
      <description>An AI coding agent runs commands with your permissions. What can go wrong, what a sandbox changes, and how OmniRush runs each project in a Linux sandbox.</description>
      <category>Sandbox</category>
      <category>Guides</category>
      <category>CLI</category>
    </item>
    <item>
      <title>Free AI coding agents: what you get, and what you really pay with</title>
      <link>https://omnirush.ai/blog/free-ai-coding-agent</link>
      <guid isPermaLink="true">https://omnirush.ai/blog/free-ai-coding-agent</guid>
      <pubDate>Mon, 19 Oct 2026 08:00:00 GMT</pubDate>
      <description>Every free AI coding agent is paid for somehow: your keys, limits, ads or data. How to tell them apart, what to check, and where OmniRush fits.</description>
      <category>Guides</category>
      <category>Models</category>
    </item>
    <item>
      <title>How to use GPT models for coding for free</title>
      <link>https://omnirush.ai/blog/use-gpt-models-free-for-coding</link>
      <guid isPermaLink="true">https://omnirush.ai/blog/use-gpt-models-free-for-coding</guid>
      <pubDate>Mon, 19 Oct 2026 08:00:00 GMT</pubDate>
      <description>Ways to code with OpenAI's GPT models without paying, what each costs you instead, and a step-by-step setup of GPT 6 Astra in OmniRush's free agent.</description>
      <category>Guides</category>
      <category>Models</category>
      <category>Walkthrough</category>
    </item>
    <item>
      <title>An AI coding agent CLI for Linux: install and first run</title>
      <link>https://omnirush.ai/blog/ai-coding-agent-cli-linux</link>
      <guid isPermaLink="true">https://omnirush.ai/blog/ai-coding-agent-cli-linux</guid>
      <pubDate>Mon, 19 Oct 2026 08:00:00 GMT</pubDate>
      <description>Install a free AI coding agent in your Linux terminal: Node.js, npm, sign-in, first task, scripts with -p, resuming chats, the sandbox and the desktop app.</description>
      <category>CLI</category>
      <category>Linux</category>
      <category>Guides</category>
    </item>
  </channel>
</rss>
