The latest articles on Forem by taekim34 (@taekim34). https://forem.com/taekim34 https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F3911526%2F05766406-74fd-4a70-aa54-8cee12ad9ebe.png https://forem.com/taekim34 en taekim34 Tue, 05 May 2026 15:36:00 +0000 https://forem.com/taekim34/we-need-a-catrun-for-the-ai-era-34a0 https://forem.com/taekim34/we-need-a-catrun-for-the-ai-era-34a0 <p><em>A 16-pixel hero in your macOS menu bar. Watches LLM traffic. That's it.</em></p> <h2> RunCat told us the CPU was busy. Nothing tells us the agent is. </h2> <p>You remember <a href="https://kyome.io/runcat/index.html?lang=en" rel="noopener noreferrer">RunCat</a> — the kitten in your menu bar that runs faster when your CPU is busy. Almost a decade old. Adorable. Useful. Asks nothing of you.</p> <p>AI-native development needs the same thing for a different signal. Not CPU. <strong>Agent traffic.</strong> Is there a live LLM request flowing right now, or is everything quiet?</p> <p>That's why I built <strong>AgentRunner</strong>.</p> <blockquote> <p>We need a CatRun for the AI era. So I made one.</p> </blockquote> <p>A 16-pixel hero in your macOS menu bar. Runs when your agent's actually working. Idle when it isn't. That's the whole UI.</p> <p><a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2F7mscgpheaiq6v6t3mxgl.gif" class="article-body-image-wrapper"><img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2F7mscgpheaiq6v6t3mxgl.gif" alt="AgentRunner demo — 23-second walkthrough" width="600" height="479"></a></p> <h2> Seven things it's built around </h2> <p><strong>1. The menu bar is where you already glance.</strong> Same place as the clock. No extra tab, no extra window, no "I'll open the dashboard later."</p> <p><strong>2. Below the noise floor.</strong> &lt;1% CPU, ~20MB RAM. Native SwiftUI. A monitor that becomes its own monitoring problem is a joke.</p> <p><strong>3. Flashy "live agent dashboards" don't last.</strong> Animated traffic, live token deltas, color-coded latency heatmaps — fun for a week, closed and forgotten by the next sprint. CatRun ran for a decade because it asked you nothing. Same spirit here.</p> <p><strong>4. Detailed analysis belongs in a different tool.</strong> Token spend, cache misses, run history — that needs report depth. That's what <a href="https://github.com/ww-w-ai/cc-token-saver" rel="noopener noreferrer">cc-token-saver</a> is for, and it gets its own post next. AgentRunner = glance. cc-token-saver = report. Don't make one app try to be both.</p> <p><strong>5. Vendor-neutral by design.</strong> It watches LLM traffic, not Claude traffic. Claude Code, Codex, Cursor, Windsurf, local LLaMA via Ollama, any agent loop hitting a model endpoint over HTTPS. No API key, no per-vendor SDK.</p> <p><strong>6. Local-only. Zero telemetry.</strong> Detection happens on your machine. The app does not phone home. No analytics SDK, no event ping. An agent monitor that ships your data anywhere doesn't deserve trust.</p> <p><strong>7. Idle vs Active. Binary.</strong> That's the entire UI. CatRun gave us a kitten that ran when CPU spiked. AgentRunner gives you a 16-pixel hero that runs when LLM traffic flows. Same spirit. Useful. Small. Invisible until you glance at it.</p> <h2> Get it </h2> <ul> <li> <strong>Repo</strong>: <a href="https://github.com/ww-w-ai/AgentRunner" rel="noopener noreferrer">https://github.com/ww-w-ai/AgentRunner</a> </li> <li> <strong>License</strong>: Apache-2.0</li> <li> <strong>Requires</strong>: macOS 13+</li> </ul> <p>cc-token-saver post: coming next.</p> showdev ai macos devtools