The angry reply, de-escalated in writing
Lesson 3 of 6 · The seller call
1 · Learn the move · Refinement loop
Angry replies come in two kinds and the first job is telling them apart. 'Stop contacting me' is an opt-out, and an opt-out is an instruction, not an opening: you honour it, log it the same day, and whether you send anything at all on that channel is a question for counsel, not the model. 'Your offer is an insult' is a person who is still talking. For that one, the first draft is always wrong in a predictable direction - defensive, or over-apologetic in a way that restarts the argument, or it slips in a re-pitch. So run the loop: draft, critique against named failure shapes, redraft. The loop is where the skill lives. Two passes usually get you something short enough to survive being forwarded to the rest of the family.
Sort these replies first: OPT-OUT (any request to stop contact) or STILL TALKING. Draft nothing for an opt-out; tell me to log it and check with counsel before any reply. For the other, write draft 1, critique it for defensiveness, over-apology, any figure, any claim and any re-pitch, then write draft 2.
2 · Your turn. You write the prompt
Two replies came in on the same Calder County mailing. A text from an unsaved number: 'STOP. How did you get this number.' And an email from Loretta Quill about parcel 14-3380-000: 'Your offer is an insult. My family worked that land for sixty years. Don't pretend you know anything about it.' Write the prompt that handles both without making either worse.
Write the prompt for this scenario as given, or swap in a real one from your own work. The score works the same either way.
Remember: the AI sees only your prompt, not this page. If the situation isn't in your prompt, it doesn't exist.
Optional. These shape the output when you run your prompt below, not your score.