This project provides a StarCraft environment for OpenAI Gym, enabling AI research within the popular real-time strategy game. It targets AI researchers and developers looking to build and test agents in a complex, dynamic environment.

How It Works

The environment is built upon Facebook's TorchCraft, acting as a bridge between the Torch deep learning framework and StarCraft. This integration allows for the control and observation of game states, facilitating the development of reinforcement learning agents.

Quick Start & Requirements

• Install OpenAI Gym and dependencies.
• Install TorchCraft and dependencies (skip Torch client).
• Install torchcraft-py and dependencies.
• Install the package: git clone https://github.com/deepcraft/gym-starcraft.git && cd gym-starcraft && pip install -e .
• Requires a running StarCraft server with BWAPI via Chaoslauncher.
• Usage example: python random_agent.py --ip $server_ip --port $server_port

Highlighted Details

• Integrates StarCraft gameplay with the OpenAI Gym API.
• Leverages TorchCraft for Torch-StarCraft communication.
• Provides example agents for quick testing.

Maintenance & Community

The project is marked "(In progress)", indicating ongoing development. No specific community links or contributor information are provided in the README.

Licensing & Compatibility

The license is not specified in the README. Compatibility for commercial use or closed-source linking is undetermined.

Limitations & Caveats

The project is explicitly stated as "In progress," suggesting potential instability or incomplete features. The README lacks details on specific StarCraft versions supported or known limitations.