Symptom
Rejection notice: "담당자 검토 결과 반려되었습니다" ("Rejected after reviewer assessment"). That is a status line, not a reason. The real reason is in the history endpoint's comment.
What it cost
Four items in the same product family → judged it a category restriction and set block flags on dozens of related items. The real reason for all of them was image specifications. Reverted every one.
★ Items registered in the same batch share the same defect. Simultaneous rejection across a product family can be coincidence.
Once more: the history query decided it
2 of several hundred rejected, right after a name-correction batch. Easy to blame the correction. History showed the same rejection reason before the correction.
★ Don't attribute a symptom right after a change to that change. History decides.