Show HN: A neural network that trains in the browser to play Clash Royale

Category: ai-ml

Tags: reinforcement-learning, game-engine, cpp, python-bindings, clash-royale, ppo, simulation, computer-vision

Score: 7.3/10 (Innovation: 7, Technical: 9, Documentation: 8, Utility: 5)

ClashRoyaleEnv is a high-performance, deterministic Clash Royale battle engine written in C++ with Python bindings, enabling reinforcement learning research on the game. It includes a full training stack with a recurrent PPO agent, lookahead search, and a perception pipeline to read the real game screen, all validated by extensive tests and detailed documentation.

Target audience: reinforcement learning researchers, game AI developers, hobbyist game players

Repository: https://itzik123.github.io/ClashRoyaleAi/lab/ · Python · MIT · 24 stars

View on Hacker News