Hello world, meet Coverage.

Coverage is a simple app, designed to make it fast and easy for anyone to capture video from different angles, simultaneously.

The idea started with my golf instructor, Jess, whose iPad is never far away when she’s coaching. She’d recently been to a golf simulator with a setup that lets you watch yourself swing from several angles at once, and she mentioned how useful that would be for coaching in general, not just in a simulator bay.

The gears started turning immediately. We all have iPhones, how hard could it be to throw something together that let them team up to act like that simulator set up?

That question turned into a weekend project, which turned into Coverage.

How it got built

I’ve been obsessed with agentic programming recently, and building Coverage felt like a perfect low stakes way to continue to learn. I’ve spent a lot of time thinking about how agentic programming can fit into a real development flow, and at this point I’m fully convinced it is the future of development.

I’m constantly wowed by just how capable agentic programming is, but haven’t been able to shake this feeling that we’re in the Wild West of tokens. Tokenmaxxing is all the rage, and it’s easy for a company the size of a small village to justify those expenses, but what about single programmers?

I’ve found that, like people, LLMs are best when they’re iterating. One agent to write the code, then another one (or two, or three) to check their work. This is all fine and well, but each of these agents is another expense, another context window you’re creating. So the question became how far can I get with less? I set myself the constraint of only using Claude’s cheapest tier, and got to work.

Yes, I definitely still hit my token limit a couple times, and spent a fair amount of time trying to minimize work whenever possible, but couldn’t be happier with the result. The biggest change for me was about picking my battles. Instead of running a full suite of reviewers on each PR, I’d have it evaluate if the changes made felt risky or not and if they were low stakes changes, just give them a cursory glance instead a full review.

That’s the whole experiment, really. See how far a little discipline and thinking ahead can get you when tokens aren’t free. Coverage is the answer - built lean, on a budget, and it works. You can check it out on the App Store here.