# New chat handoff summary prompt: free ChatGPT template > A copy-ready prompt that boils a long ChatGPT chat down to a fact sheet, so you can start a new chat without re-pasting or carrying wrong turns. Source: https://learn.ynteractive.com/content/prompts/new-chat-handoff-summary-prompt · Updated 2026-10-16 · Free from Stepthrough (https://learn.ynteractive.com) [SOPs and operations](https://learn.ynteractive.com/content/prompts#job-sops-and-operations) · [Operations](https://learn.ynteractive.com/content/prompts#industry-operations) · Works in ChatGPT, Claude, Gemini and Copilot Updated October 16, 2026 · By [Robert Breen](https://learn.ynteractive.com/content/about) Use this when a chat has run long, got confused or taken a wrong turn, and you want a clean start without losing the facts you already settled. You ask the AI to write the settled facts down, then paste that short sheet into a new chat. Read the sheet before you reuse it. It is a summary, so it can drop or change a detail. Compare it with your own notes, and leave out names and numbers the next task does not need. ## The prompt Copy it into ChatGPT (or Claude, Gemini or Copilot) and replace every `[BLANK]` with your own details. It uses the four parts from [Prompt Writing 101](https://learn.ynteractive.com/content/prompt-writing-101-role-context-task-format): Role, Context, Task and Format. Copy prompt ```text Role: You are an assistant helping an office manager hand over a working chat about [TOPIC] to a fresh chat. Context: This chat has [WHAT HAPPENED]. I want to start a new chat for the same job and paste in only what is settled. Task: List only the facts we have agreed on in this chat, each with its source if I gave one. Put anything uncertain or contradicted under a separate heading called Needs checking. Do not add facts I did not give you, and do not include wrong turns or rejected drafts. Format: A short sheet with two headings: Settled facts (a numbered list) and Needs checking (a list). Plain words, under 200 words in total. ``` ## Fill in the blanks **`[TOPIC]`** The job the chat was about, for example a customer quote. **`[WHAT HAPPENED]`** A few words on why you are restarting, such as changes in the headcount and one draft that went the wrong way. ## Example, filled in **A made-up example** from the free lesson [Stop Losing Context: When to Start a New Chat vs Keep Going using ChatGPT](https://learn.ynteractive.com/content/chatgpt-stop-losing-context-new-chat-vs-keep-going). Oakwood Library is the made-up customer in the Stop Losing Context lesson. The details below are an illustration of how an office manager might fill the prompt in, not the lesson's exact data. Copy example ```text Role: You are an assistant helping an office manager hand over a working chat about the staff breakfast quote for Oakwood Library to a fresh chat. Context: This chat has a quote, a headcount that changed from 80 to 100 guests and a text message to Sam. I want to start a new chat for the same job and paste in only what is settled. Task: List only the facts we have agreed on in this chat, each with its source if I gave one. Put anything uncertain or contradicted under a separate heading called Needs checking. Do not add facts I did not give you, and do not include wrong turns or rejected drafts. Format: A short sheet with two headings: Settled facts (a numbered list) and Needs checking (a list). Plain words, under 200 words in total. ``` ## What a good answer looks like - Only facts you gave appear, each tied to where it came from. - Anything unsure is parked under Needs checking instead of being stated as true. - The sheet is short enough to paste into a new chat in one go. ## How to check it before you use it An AI draft can sound right and still say something you never told it. Use the habit from the [Reply Faster lesson](https://learn.ynteractive.com/content/chatgpt-reply-to-customer-emails): check every promise, name, date and number against what you gave it before anything goes out. - Compare every number and date with your fact sheet before you paste the summary anywhere. - Delete names, account numbers and anything else the next task does not need. - Read the Needs checking list first. That is where the old chat's mistakes hide. - If the sheet is long, the old chat was probably several jobs. Split them. ## Why this prompt works A long chat carries everything in it into the next answer, including a wrong turn from twenty messages ago. A short, checked sheet gives the new chat only what is true. Asking for a separate Needs checking heading means the AI has somewhere honest to put its doubts. ## Practice it in the free lesson **[Stop Losing Context: When to Start a New Chat vs Keep Going using ChatGPT](https://learn.ynteractive.com/content/chatgpt-stop-losing-context-new-chat-vs-keep-going)**: Quote a customer in one ChatGPT chat, change the numbers without re-pasting anything, then start a clean chat for the next customer so their details never mix. You do every step yourself in a practice copy of ChatGPT, and nothing touches your real accounts. [Start the free lesson](https://learn.ynteractive.com/modules/chatgpt-stop-losing-context) Also useful: [Prompt Writing 101: Role, Context, Task, Format](https://learn.ynteractive.com/content/prompt-writing-101-role-context-task-format) ## Terms used here [New chat](https://learn.ynteractive.com/content/glossary/new-chat) · [Follow-up prompt](https://learn.ynteractive.com/content/glossary/follow-up-prompt) · [Context window](https://learn.ynteractive.com/content/glossary/context-window) ## Related prompts - [Meeting transcript cleanup prompt](https://learn.ynteractive.com/content/prompts/meeting-transcript-cleanup-prompt): A copy-ready prompt that cleans a raw meeting transcript before you summarize it: fixes speaker labels, removes filler, flags… - [Email thread summary prompt](https://learn.ynteractive.com/content/prompts/email-thread-summary-prompt): A ChatGPT prompt that turns a long email thread into a short handoff: the situation, what was promised, each question marked… - [Decision log prompt](https://learn.ynteractive.com/content/prompts/decision-log-prompt): Pull every settled decision out of meeting notes and messages into a dated log: who decided, why, and when to revisit, plus what… ## More sops and operations prompts - [New matter client document checklist prompt](https://learn.ynteractive.com/content/prompts/client-document-checklist-for-new-matter-prompt): Build a plain-language list of documents a new client should gather for their matter, with a short reason for each and a way to… - [Policy summary for staff prompt](https://learn.ynteractive.com/content/prompts/policy-summary-for-staff-prompt): Turn a long internal policy into a plain staff summary: who it applies to, what to do, limits and deadlines, with unclear lines… - [Post-job care instructions prompt](https://learn.ynteractive.com/content/prompts/post-job-care-instructions-prompt): Turn your own care notes for one job type into a leave-behind sheet: what's normal, simple upkeep, what not to touch, and when… ## Visual guide This prompt as a quick visual guide you can save or share, with the same text written out. - [Mistake: One ChatGPT Chat for Every Customer](https://learn.ynteractive.com/content/visual/mistake-one-chat-for-every-customer): A visual guide to a common ChatGPT mistake: quoting a second customer in the first customer's chat. Start a new chat for each customer. [All visual guides](https://learn.ynteractive.com/content/visual) ## Frequently asked questions **Why not just keep going in the long chat?** Long chats carry old details and wrong turns into new answers. A fresh chat with a checked fact sheet is cleaner, as the free lesson shows when it starts a new customer in a new chat. **Does the new chat remember the old one?** Not by itself. It only knows what you paste, plus anything saved in memory if that is turned on in your settings. [All free prompt templates](https://learn.ynteractive.com/content/prompts) · [AI glossary](https://learn.ynteractive.com/content/glossary)