To stress-test it and see what people actually build, we're running a competition:

- Pay $35 to enter - You get a limited number of credits to build something - Submit it to a public leaderboard - Community votes on the best submissions - $50K first place, $10K second, $5K third

Three winners. That's it. The constraint is intentional. We wanted to see what's possible when you can only describe what you want in a few shots. Some of the early submissions have been genuinely surprising — real, usable apps built by people who've never written code.

The platform is live and the leaderboard already has submissions. Would love HN's feedback on:

1. The platform itself — what works, what breaks, what you'd improve 2. Whether this kind of constraint-based competition surfaces interesting use cases 3. What you'd build if you only had a few prompts

Happy to answer any technical questions about the architecture.