task_games__cards__trick_taking_winner_label

Contract

1. Domain: games
2. Scene id: cards
3. Public task id: task_games__cards__trick_taking_winner_label
4. Supported query_id values: single
5. Answer schema: string_label
6. Annotation schema: bbox
7. Program schema: label(arg_extreme(played_cards, metric=trick_order_metric(card, lead_suit, trump_suit), direction=winning)); scene=cards; scope=trick_taking_winner_label

Program Contract

Program: label(arg_extreme(played_cards, metric=trick_order_metric(card, lead_suit, trump_suit), direction=winning)); scene=cards; scope=trick_taking_winner_label

Candidate set: the visible game board, pieces, tokens, cards, tiles, marked state, legal-move cues, result panels, and labeled options inside the trick_taking_winner_label objective scope. Operands: visible scene state and prompt-bound operands named by arg_extreme, played_cards, metric, trick_order_metric, card, lead_suit, trump_suit, direction, winning, cards, trick_taking_winner_label. Operation: evaluate label 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 bbox schema; the prompt/annotation contract defines the minimal visual witnesses. Query ids: single.

Reasoning Operations

Families: filtering, comparison, ranking

Generation Notes

1. Prompt wording comes from src/trace_tasks/resources/prompts/games/cards/games_cards_v1.json.
2. Annotation is projected from the same generated game state used for answer verification.