Same chat shell as Sandbox B mobile. MCQ cards stack full-width. Suggestion chips horizontal-scroll. Same sticky composer pattern. Two frames: Q4 confidence (mid-flow) and Q12 score reveal (the climax — adapted for mobile).
Marketing Director · B2B SaaS · 15 hrs/wk on writing & analysis · uses ChatGPT + Claude
Honest one. When you use AI at work, how do you actually feel?
No wrong answer. Anxious is welcome.
Twelve questions. Twelve honest answers. A clear read on where you are and where this goes.
You understand key concepts and you're ready to level up to strategic, workflow-grounded application.
You're ahead of where you think you are — and stalled in a fixable place.
You ship fast but verify every output. Not a confidence problem — a structure problem.
No "what is an LLM." Start at MetaPrompting.
Less reading. More attempts with feedback.
Maps to your launch playbook work.
Same chat shell as Sandbox B mobile. The score-reveal moment uses a sticky bottom CTA so the primary action is always thumb-reachable, regardless of how far the user scrolls through their report.