task_games__connect_four__winning_move_column_label

Contract

1. Domain: games
2. Scene id: connect_four
3. Public task id: task_games__connect_four__winning_move_column_label
4. Supported query_id values: single
5. Answer schema: label_string
6. Annotation schema: point
7. Program schema: select(column_label, legal_drop_result=immediate_win_for_current_player); scene=connect_four; scope=winning_move_column_label

Program Contract

Program: select(column_label, legal_drop_result=immediate_win_for_current_player); scene=connect_four; scope=winning_move_column_label

Candidate set: the visible game board, pieces, tokens, cards, tiles, marked state, legal-move cues, result panels, and labeled options inside the winning_move_column_label objective scope. Operands: visible scene state and prompt-bound operands named by column_label, legal_drop_result, immediate_win_for_current_player, connect_four, winning_move_column_label. Operation: evaluate select over the candidate set using the visible game state, rules, legal moves, comparisons, counts, simulations, or option-selection constraints encoded in the program expression; generation enforces a unique final answer. Output binding: answer uses the string schema; generation binds a unique final answer. Annotation witnesses: annotation uses the point schema; the prompt/annotation contract defines the minimal visual witnesses. Query ids: single.

Reasoning Operations

Families: state_update

Generation Notes

1. Column labels are rendered below the board; annotation marks the center of the selected column's landing cell, not the label text.
2. The answer landing cell is not visibly highlighted in the rendered image.
3. Annotation is projected from the same generated game state used for answer verification.