task_pages__infographic__global_metric_ranked_item_label

Identity

1. Domain: pages
2. Scene id: infographic
3. Source path: src/trace_tasks/tasks/pages/infographic/global_metric_ranked_item_label.py
4. Task id: task_pages__infographic__global_metric_ranked_item_label

Program Contract

1. Program schema: selection.ranked_item(candidate_set=all_metric_cards, metric=printed_value, rank_direction=highest|lowest); scene=infographic; scope=one multi-section metric-card infographic
2. Contract: rank every visible metric card in the infographic by printed value, then return the label of the card at the requested rank.
3. Public query ids: nth_highest_metric_label, nth_lowest_metric_label
4. Answer schema: string
5. Annotation schema: scalar bbox around the selected metric card.
6. Query argument axes: rank direction and rank position.
7. scalar_annotation_checked=true

Reasoning Operations

Families: ranking

Prompt + Trace

1. Prompt bundle: pages_infographic_v1
2. Scene key: infographic_metric_arithmetic
3. Task key: metric_arithmetic_query
4. Trace records ranked candidates, rank direction, rank position, selected label/value, and the selected metric-card bbox.
5. Generation guarantees unique printed values in the ranked candidate set.