metadata: migrate score_type -> score_pipeline (atomic stage contract; see mm-eval scorer docs/en/SCORING.md)

Update metadata.json

Push fill_in_blank config

Push choose_txt config

Push choose_img config

metadata: 2026-07-07 revised taxonomy (task_type format vocabulary multiple_choice/yes_no/short_answer/free_form; score_type incl. none; re-verified protocol evidence; repaired prompt provenance)

metadata: add score_type/score_params/score_protocol (official grading protocol taxonomy); task_type is now format-only (judge_scored migrated to vqa + llm_judge). See dataset-upload skill references/metadata-json.md.

Tighten prompt_template_source.notes: explicit 5-tier classification per 2026-06-01 priority rules

Refine prompt_template_source: cite verifiable official sources per 2026-05-30 deep audit

Populate prompt_template_source for all subsets per 2026-05-30 audit

Restore metadata.json

Restructure: per-task config fill_in_blank

Restructure: per-task config choose_txt

Restructure: per-task config choose_img

initial commit