task_games__bubble_shooter__pop_target_label

Contract

1. Domain: games
2. Scene id: bubble_shooter
3. Public task id: task_games__bubble_shooter__pop_target_label
4. Supported query_id values: single
5. Answer schema: string_label
6. Annotation schema: bbox
7. Program schema: label(unique_landing_target_where(count(existing_same_color_component_adjacent_to(target, shooter_color)) + 1 >= 3)); scene=bubble_shooter; scope=pop_target_label

Program Contract

Program: label(unique_landing_target_where(count(existing_same_color_component_adjacent_to(target, shooter_color)) + 1 >= 3)); scene=bubble_shooter; scope=pop_target_label

Candidate set: the visible game board, pieces, tokens, cards, tiles, marked state, legal-move cues, result panels, and labeled options inside the pop_target_label objective scope. Operands: visible scene state and prompt-bound operands named by unique_landing_target_where, existing_same_color_component_adjacent_to, shooter_color, bubble_shooter, pop_target_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, counting, comparison, topology, state_update

Generation Notes

1. This task is owned by the current-layout public file src/trace_tasks/tasks/games/bubble_shooter/pop_target_label.py.
2. The public task id selects the objective; query_id is single.
3. Annotation is projected from the same generated game state used for answer verification.