| # Copyright (c) Meta Platforms, Inc. and affiliates. | |
| # All rights reserved. | |
| # | |
| # This source code is licensed under the BSD-style license found in the | |
| # LICENSE file in the root directory of this source tree. | |
| """ | |
| Chess Environment for OpenEnv. | |
| This module provides OpenEnv integration for chess, using the moonfish | |
| chess engine for position evaluation and opponent play. | |
| Example: | |
| >>> from envs.chess_env import ChessEnv, ChessAction | |
| >>> | |
| >>> # Connect to a running server or start via Docker | |
| >>> env = ChessEnv.from_docker_image("chess-env:latest") | |
| >>> | |
| >>> # Reset and interact | |
| >>> result = env.reset() | |
| >>> print(result.observation.fen) | |
| >>> print(result.observation.legal_moves) | |
| >>> | |
| >>> result = env.step(ChessAction(move="e2e4")) | |
| >>> print(result.reward, result.done) | |
| >>> | |
| >>> # Cleanup | |
| >>> env.close() | |
| """ | |
| from .client import ChessEnv | |
| from .models import ChessAction, ChessObservation, ChessState | |
| __all__ = ["ChessEnv", "ChessAction", "ChessObservation", "ChessState"] | |