Queen Chess

Tools

• queen_pawn_8_analyse_position Ask Queen, a chess language model, for its best move in a position, with its evaluation and explanation. Queen plays at roughly 2700 Lichess Elo; its evaluations are generated estimates, not engine search results, and its prose can misdescribe tactics. One call takes up to a minute. Returns: {"analysis": Markdown explanation, "best_move_san", "best_move_uci", "evaluation", "critical_line_san",, "critical_line_uci", "fen": the position analysed, ...}
• queen_pawn_8_annotate_game Annotate one game of a PGN with Queen's comments, as the Annotate tab does. With [%eval] comments in the PGN, Queen analyses only the moves those evals flag; without them, it screens each move in range and comments where it would have played differently. Each move takes up to a minute: from Python, use client.submit() and read the updates as they arrive; over MCP only the last update is returned. A long call can be stopped by ZeroGPU (its login token expires after a few minutes); the last update then says where it stopped and still carries the PGN annotated so far, so keep max_moves small. Returns: After each analysed move, {"status": str, "comments": [Markdown str], "pgn": None};, the last update has "pgn" set to the annotated PGN.

No tool matches.

As the server's own listing gives them. We never call a tool.