An AI that makes you
think first

Most AI is built to do your thinking for you. This brief does the opposite. Paste it into a new Claude conversation and it asks you eight questions about your working year, makes you place yourself before it places you, and refuses to say a word about you until you have done the work. Then it gives you the read your manager will not and your friends cannot.

Someone describes their year in five stories and says their strength is communication. Asked what repeats across the five, they cannot see it. In every one, the part that needed them to walk into a room or send the thing was the part that moved to somebody else. Nobody was going to say that to them. They had to be made to look.

Three years of AI have trained people to hand over the thinking and take the answer. That is the wrong shape for this job, because the answer only means something if you did the looking. Claude runs the interview one question at a time, asks you what repeats before it tells you, and treats the gap between where you placed yourself and where your own stories place you as the finding. You come out having thought harder about your year than anyone has asked you to, and with one thing to do about it.

• The brief itself, in one block you copy or download, so what you hand Claude is exactly what you have read.
• The eight questions it asks before it says anything about you, including where you think you are, what you avoided this year, and what it does when your answer blames somebody else.
• The guardrails it holds itself to, including that it never grades you, never answers a question you could answer yourself, and never writes the version for your manager.

Someone describes their year: three projects, one that went badly, the praise they got, the conversation they put off. Asked where they think they are, they say on track. Asked what repeats across their own stories, they name the wrong thing. The run then puts one question: in all three projects, the part handed to somebody else was the follow up, and the conversation put off was also a follow up. Is that one thing or three? The gap between on track and that question is the read.