Enables an LLM to play the classic roguelike game NetHack.
This tool provides a bridge for large language models to interact with and play NetHack, a complex and challenging roguelike game. Leveraging the opencode agent environment, it facilitates an AI's autonomous navigation and decision-making within the game. By managing the NetHack session within a tmux environment, it offers a robust setup for developing and testing intelligent agents in dynamic, text-based gaming scenarios.
Key Features
01LLM-driven NetHack gameplay via opencode agent environment
02Modular Go-based architecture for core logic
030 GitHub stars
04Seamless game session management using tmux
05Streamlined build and test processes with Makefile
Use Cases
01Develop and test AI agents in complex game environments
02Explore autonomous gameplay strategies for roguelike games
03Experiment with LLM capabilities in dynamic, unpredictable scenarios
