games¶

Haive Games - Comprehensive game environment collection.

This package provides a rich ecosystem of game implementations designed for AI agent training, research, and entertainment. Games range from classic board games to card games to strategy games.

Game Categories:

  • Board Games: Chess, Checkers, Go, Reversi, Connect4, Tic-Tac-Toe

  • Card Games: Poker, Hold’em, Blackjack, UNO

  • Strategy Games: Risk, Monopoly, Debate, Mafia

  • Puzzle Games: Mancala, Mastermind, Nim, Sudoku, Wordle

  • Social Games: Among Us, Debate, Clue

Each game includes:

  • AI agent implementations with configurable difficulty

  • State management and game rules enforcement

  • Rich UI components for visualization

  • Tournament and benchmarking capabilities

  • Extensible configuration systems

Examples

Quick tic-tac-toe game:

from haive.games.tic_tac_toe import TicTacToeAgent, TicTacToeState

agent = TicTacToeAgent()
game_state = TicTacToeState()
result = agent.play(game_state)

Chess with custom agents:

from haive.games.chess import ChessAgent, ChessState

white_agent = ChessAgent(name="white", difficulty="expert")
black_agent = ChessAgent(name="black", difficulty="intermediate")

Multi-player poker:

from haive.games.poker import PokerAgent, PokerState

agents = [PokerAgent(f"player_{i}") for i in range(4)]
game = PokerState(players=agents)

All games support both human and AI players, with extensive configuration options for gameplay variants, AI behavior, and visualization preferences.

Submodules¶