I live in San Francisco. People visit. People ask where to eat. I love this city and I love food so I always have opinions, but at some point I realized I was copy-pasting the same list of spots to different group chats every other week.
So I built a website for it.
sfcravings.channel is a curated list of my favorite places to eat in SF, broken into three categories: comfort foods, sunny weather spots, and nighttime spots. That's it. No reviews, no star ratings, no "sort by cuisine." Just my picks, organized by vibe.
I styled the whole thing like a 90s Weather Channel broadcast. Vaporwave soundtrack included. Because if you're going to build something just for yourself and your friends, you might as well have fun with it.
Why I'm writing about this
I've been building small personal projects with Claude Code for a few months now. A transit tracker configured for my specific commute. A metal show aggregator for SF venues. An offline survival knowledge base. Each one started the same way: I had a tiny problem that only I cared about, and I realized I could just... build a thing for it.
This is the part that still feels new to me. A year ago I would have opened a Google Doc and typed out my restaurant list. Or made a shared note. Building a whole website for something this small would have felt absurd. But the effort to go from "I want a page" to "I have a page" has collapsed. Building personal software is now about as hard as putting together a slide deck. Maybe easier, because no one's going to ask for revisions.
Personal software
I think there's something worth paying attention to here. We're at a point where anyone with a clear idea of what they want can build software for an audience of one. Not an app. Not a startup. Just a tiny piece of software that solves a problem in your specific life, built exactly the way you want it.
The restaurant site took me about 10 hours total. Some of that was messing around with the Weather Channel aesthetic and getting the soundtrack right. The actual functional part of the site, the list of restaurants organized by category, that was fast. Most of the time went into making it weird and fun, which is kind of the whole point.
I don't think every project needs to be a portfolio piece or a SaaS product or something you'd pitch to investors. Some of the best things you can build are small, dumb, and only useful to you and maybe a dozen people you know. The tools are good enough now that the cost of building something personal is basically just your time and your taste.
The site
If you're visiting SF or just looking for a new spot, check it out: sfcravings.channel
Turn your sound on.