We’re releasing an open benchmark for evaluating AI coding agents on real-world Kotlin tasks. Compare agents, inspect the methodology, and see the results on the leaderboard.