task_games__platformer__jump_landing_label

Contract

1. Domain: games
2. Scene id: platformer
3. Public task id: task_games__platformer__jump_landing_label
4. Supported query_id values: single
5. Answer schema: string_label
6. Annotation schema: bbox
7. Program schema: label(landing_platform(marked_jump)); scene=platformer; scope=jump_landing_label

Program Contract

Program: scene=platformer; scope=jump_landing_label; program=label(landing_platform(shown_jump_arc))

Candidate set: the visible game board, pieces, tokens, cards, tiles, marked state, legal-move cues, result panels, and labeled options inside the jump_landing_label objective scope. Operands: visible scene state and prompt-bound operands named by platformer, jump_landing_label, program, label, landing_platform, shown_jump_arc. Operation: evaluate scene=platformer 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: spatial_relations

Generation Notes

1. Query ids are internal replay/sampling keys and do not define public task units.
2. Annotation is projected from the same generated game state used for answer verification.