REVISION - Client Feedback Kit | PROMPT v2 | Solo AI Content (soloaicontent.com) HOW TO USE 1. Copy everything between "START OF PROMPT" and "END OF PROMPT" into your AI chat. 2. Replace the two [PASTE ...] parts at the bottom with your own text. 3. Read every row of the answer. Check the exact words and the scope before you act. BEFORE YOU PASTE Remove or replace names, emails, phone numbers, prices, passwords and anything confidential that you do not need. Only use an AI tool that you are allowed to use with this information. More in USAGE.txt. === START OF PROMPT === You are helping me sort out messy client feedback. Use ONLY the text I paste. Never invent a decision, an approval, a date, a person or a quote. Start your answer with one sentence saying this is a draft for me to check before I act. For every distinct change request (ignore thanks, small talk, invoices and my own replies): - List each message that mentions it: SOURCE (Slack / Email / Call notes, who, when - use the date or day I gave; if it is missing, write "not given") and their EXACT WORDS, copied without rewording. - STATUS: CLEAR (the messages agree, or there is one clear instruction) / CONFLICT (messages disagree) / UNCLEAR (a question, "maybe", "if it's quick", "unless...", "whatever you think", "decide later", or a vague "sounds good") / CHECK SCOPE (looks like extra work beyond the ORIGINAL SCOPE below). - CAN WE START? Answer Yes only if ALL of these are true: the status is CLEAR, the request is inside the original scope, and the text I gave you shows who decided it. If the scope or the person with authority cannot be determined from my text, answer No and say what is missing. If unsure, No. - For anything that is not a clear Yes, write one short question for the client that quotes what was said. Rules: The newest message does NOT automatically win. A postponed decision ("decide later", "let's see it without and decide") is NOT a decision: it does not settle a conflict. "X is fine with it", said by someone else, is unconfirmed. If you can't tell who has the authority to decide, say so. Don't decide anything for me; if you're unsure, say so. Give me: A) A CSV table in a code block, with exactly these six columns, in this order, comma-separated: "REQUEST","SOURCE","EXACT WORDS","STATUS","CAN WE START?","QUESTION TO CLIENT". Print the header line once. Put every field in double quotes. If a field contains a double quote, double it (""). Never put a line break inside a field. One row per message; repeat the request, status, can-we-start and question on every row of the same request. B) START NOW: the requests I can start today. C) WAIT: the requests that need a client answer first. D) A short, friendly message to the client with only the questions. ORIGINAL SCOPE: [PASTE what was agreed: pages/deliverables, revision rounds included, which round this is, and who can approve changes on the client side] CLIENT FEEDBACK (label each part Slack / Email / Call notes, with the date): [PASTE the messages] === END OF PROMPT ===