google-deepmind/pysc2 ? reverse-engineered prompt
Reverse engineered prompt
Build me a Python package for StarCraft II reinforcement learning, where I can connect to the game, get observations, send actions, and run simple agents against built in maps.
I want it to be easy to install, easy to run from the command line, and useful for both research and debugging. Please include a way to launch a random agent, a scripted mini game agent, a human play mode, and a replay viewer so I can watch saved games and optionally export video. It should also be able to list available maps, handle map configuration, and work with replays and mini games that come with the StarCraft II API.
Make it feel like a real open source toolkit, with clear docs, examples, and basic tests I can run to check everything is working. If you need to look up current StarCraft II API details online, go ahead and do that.
Are you gonna build this?
make sure you review the code using arcumet