<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>hsynneva notebook</title>
    <link>https://hsynneva.com/</link>
    <description>AI experiments, automation workflows, FPV atmosphere, and practical systems notes.</description>
    <language>en</language>
    <atom:link href="https://hsynneva.com/rss.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>A safer lane for AI code triage</title>
      <link>https://hsynneva.com/blog-clawsweeper-operator-ledger.html</link>
      <guid>https://hsynneva.com/blog-clawsweeper-operator-ledger.html</guid>
      <pubDate>Mon, 27 Apr 2026 00:00:00 +0000</pubDate>
      <description>A product note on turning ClawSweeper's useful pattern into a local operator ledger for AI-assisted issue and pull request triage.</description>
    </item>
    <item>
      <title>AI ops on an early-access game server</title>
      <link>https://hsynneva.com/blog-windrose-game-server-ai-ops.html</link>
      <guid>https://hsynneva.com/blog-windrose-game-server-ai-ops.html</guid>
      <pubDate>Sun, 26 Apr 2026 00:00:00 +0000</pubDate>
      <description>A Windrose server went live on a Windows 11 VM the same hour a friend asked, then stabilized through practical AI-assisted operations.</description>
    </item>
    <item>
      <title>The operator control plane is the product</title>
      <link>https://hsynneva.com/blog-operator-control-plane.html</link>
      <guid>https://hsynneva.com/blog-operator-control-plane.html</guid>
      <pubDate>Fri, 24 Apr 2026 00:00:00 +0000</pubDate>
      <description>A practical note on why the useful AI stack is not a model choice but the control surface around delegation, verification, memory, and deployment.</description>
    </item>
    <item>
      <title>AI design tools need an apply boundary</title>
      <link>https://hsynneva.com/blog-ai-design-apply-boundary.html</link>
      <guid>https://hsynneva.com/blog-ai-design-apply-boundary.html</guid>
      <pubDate>Wed, 22 Apr 2026 00:00:00 +0000</pubDate>
      <description>A practical note on using AI for frontend design without letting generated patches silently become production changes.</description>
    </item>
    <item>
      <title>Local-first memory for agents is an operations problem</title>
      <link>https://hsynneva.com/blog-local-first-agent-memory.html</link>
      <guid>https://hsynneva.com/blog-local-first-agent-memory.html</guid>
      <pubDate>Sun, 19 Apr 2026 00:00:00 +0000</pubDate>
      <description>Persistent memory helps only when it reduces repeated steering without becoming a second, fuzzier source of truth.</description>
    </item>
    <item>
      <title>Boring website ops beats heroic redesigns</title>
      <link>https://hsynneva.com/blog-boring-website-ops.html</link>
      <guid>https://hsynneva.com/blog-boring-website-ops.html</guid>
      <pubDate>Thu, 16 Apr 2026 00:00:00 +0000</pubDate>
      <description>The useful web operations loop is small: check live status, crawlability, deployment freshness, screenshots, and the few regressions users actually feel.</description>
    </item>
    <item>
      <title>FPV drone maintenance 101: keeping a fleet airworthy</title>
      <link>https://hsynneva.com/blog-fpv-drone-maintenance.html</link>
      <guid>https://hsynneva.com/blog-fpv-drone-maintenance.html</guid>
      <pubDate>Sun, 05 Apr 2026 00:00:00 +0000</pubDate>
      <description>Essential checks, common repairs, and preventive maintenance to keep an FPV fleet airworthy.</description>
    </item>
    <item>
      <title>I built a multi-agent AI system that actually works</title>
      <link>https://hsynneva.com/blog-multi-agent-ai.html</link>
      <guid>https://hsynneva.com/blog-multi-agent-ai.html</guid>
      <pubDate>Fri, 20 Mar 2026 00:00:00 +0000</pubDate>
      <description>A layered multi-agent system with 8 agents, local LLMs on Ollama, and Claude as orchestrator.</description>
    </item>
  </channel>
</rss>
