task_games__marble_chain__max_pop_direction_label

Public taxonomy: games -> marble_chain -> task_games__marble_chain__max_pop_direction_label.

Contract

1. Domain: games
2. Scene id: marble_chain
3. Public task id: task_games__marble_chain__max_pop_direction_label
4. Supported query_id values: single
5. Answer schema: option_letter
6. Annotation schema: point
7. Program schema: argmax(labelled_shot_options, metric=immediate_existing_marble_pop_count); scene=marble_chain; scope=max_pop_direction_label

Program Contract

Program: argmax(labelled_shot_options, metric=immediate_existing_marble_pop_count); scene=marble_chain; scope=max_pop_direction_label

Candidate set: the visible game board, pieces, tokens, cards, tiles, marked state, legal-move cues, result panels, and labeled options inside the max_pop_direction_label objective scope. Operands: visible scene state and prompt-bound operands named by labelled_shot_options, metric, immediate_existing_marble_pop_count, marble_chain, max_pop_direction_label. Operation: evaluate argmax 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 option_letter 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: counting, ranking, topology, state_update

Generator

• Implementation: src/trace_tasks/tasks/games/marble_chain/max_pop_direction_label.py
• Config: src/trace_tasks/resources/configs/domains/games/marble_chain.yaml
• Prompt bundle: src/trace_tasks/resources/prompts/games/marble_chain/games_marble_chain_v1.json