Hi Product Hunt! 👋

I'm Philip, CTO & co-founder of ThinkCode — a three-person team from Seoul (CEO, PM, and me on code). We build coding education infrastructure: browser-based coding containers, live sessions, replays. It currently powers a 480-hour AI bootcamp in Korea. Along the way we noticed something odd — everyone codes with AI now, but every practice platform treats AI as cheating. So we built the opposite: a zero-setup playground where you sign in with your own Claude/ChatGPT subscription and solve problems designed for the AI era — messy server logs, ledgers with injected bugs, tasks where directing and verifying your AI is the skill. We're in public beta and launching here soon. What should coding practice look like when AI is a given? Genuinely curious what this community thinks.