How Much Review Is Too Much Review?
A human reviewer ends: they read the diff, satisfy themselves the change is safe, and ship it. An AI told to 'find issues' never reaches that point — finding is the goal, so it …
Posts
Apple's Device Mockups Are All the Wrong Size
Apple's official device design assets each render at their device's native screen resolution, so an iPhone PNG stands as tall as a MacBook and nothing lines up when you composite …
Buy AppleCare for a Family Member's Device Without Handing Them Your Credit Card
Apple sells AppleCare as if the buyer and the device's owner are always the same person. They're not. Buying coverage for a kid's iPad or a parent's iPhone seems to demand putting …
Two Gates Guard Unattended macOS UI Tests
Getting XCUITest to run unattended on a headless Mac mini looked like one permission problem. It was two: TCC privacy grants needed an MDM-delivered PPPC profile, while Automation …
Your Smartest AI Is a Lousy Junior Engineer
I used to treat model choice like a volume knob: always turn it up, always use the smartest model. Hundreds of hours of agentic coding later, I've learned the strongest model is …
Catching Twilio SMS Locally: MessagePit Extends Mailpit
Mailpit solved local email testing years ago. SMS testing never got the same treatment — most teams either disable SMS in dev or eat the Twilio bill. MessagePit is a fork of Ralph …
Your HTML Is Lying to AI Agents
Most personal sites have two versions: the rendered page humans see, and the raw HTML machines read. CSS flexbox makes the sidebar appear left and the article appear right — but …
Free Doesn't Mean Open: How AI Is Unbundling the Open Source Bargain
For decades, 'free software' and 'open source' were nearly synonymous — because building something worth sharing required a community. AI has dissolved that constraint. But the …
A Content Pipeline with Critique Loops and Editorial Gates
This blog runs on a custom automation pipeline where AI agents compete, critique, and synthesize every post. Here's how the /create-blog-post command, a DALL-E 3 image generation …
The Rise of the Agent Wrangler
AI agents can write code, fix bugs, and ship features — but who can be trusted to ship production software when most implementation is delegated? Meet the Agent Wrangler: the …