For editorial images, prompt values are applied in this order:
1) Subject/Location/Actions fields (if filled),
2) checked angles (if selected),
3) default topic-based scene logic.
Target Term vs Headline vs Topic
Target term chooses which term record to save under.
Headline is the text context used for the image prompt (and optional overlay).
Topic steers scene category and icon style when no manual override is provided.
Example
Create Image (target term): "Do You Have To Put LLC in Your Business Name?"
Headline: "Alaska Seat Belt Laws"
Topic: "vehicle law"
Subject: "Latino man in his 40s"
Location: "driver seat of a modern car in Anchorage"
Actions: "fastening seat belt and checking mirrors"
Result: the generated prompt uses the Subject/Location/Actions above, while Topic still guides legal icon set and styling.
