sokobot and sokoban-AI
These are competitors offering overlapping Sokoban AI implementations using comparable search strategies (BFS, DFS, A*) and heuristics, with the primary differentiator being that angusfung/sokoban-AI includes deadlock detection while rumbleFTW/sokobot provides a complete playable game interface.
About sokobot
rumbleFTW/sokobot
Full-Fledged Sokoban game with AI using Depth-First, Breadth-First strategies and A* Pathfinding with Manhattan heuristic
About sokoban-AI
angusfung/sokoban-AI
AI of the game Sokoban using search strategies and heuristics with deadlock checking.
Implements multiple search algorithms (DFS, BFS, A*, uniform-cost, anytime variants) with interchangeable heuristics (Manhattan distance, Euclidean, displaced blocks) and memoization via dictionary lookup to optimize performance. The modular architecture allows custom search strategies and heuristics to be added directly in solution.py, enabling experimentation with advanced techniques. Test cases validate performance across different solver configurations.
Scores updated daily from GitHub, PyPI, and npm data. How scores work