Making Danger Obvious: Feedback Wanted on My AI Response Footer (Risk + Confidence UI)