Working with AI Conversations

Turn a Good Chat Into a Reusable Prompt

Studies a conversation that went well and writes a reusable prompt template with variables, so you can repeat the result next time.

Designed for general-purpose AI assistantsPublished November 2026

Customize

Fill in the fields and the prompt updates. What you type stays in this page and is never saved or sent.

Preview

This conversation produced a result I want to get again with different inputs. Turn it into a reusable prompt.

What the successful result was: {{what worked}}
What will change from one use to the next: the topic and the source material
Where I will use it: any AI assistant

Work in this order:
1. Find the turning points: which of my messages, corrections or constraints made the biggest difference to the quality of your answers? Quote them.
2. Separate what stayed the same every time (the task, the format, the rules, the audience) from what will vary (the inputs).
3. Write the prompt as a single message I can paste into a fresh conversation. Use variables in double curly braces for everything that varies, and give a sensible default for a variable only where one exists. Put an instruction at the top that tells the assistant to ask me for anything essential that is missing instead of guessing.
4. Add an output format section that reproduces the structure of the best answer in this conversation.
5. Add a short list of rules, taken only from corrections I actually made here, for example things I told you to stop doing.
6. List the assumptions the prompt makes that a new reader might not share, and one example of a good input.

Rules: do not add requirements I never expressed, and do not include details specific to this conversation, such as names or numbers, unless they are meant to be variables. Keep the finished prompt under 350 words. If a turning point is unclear, ask me before writing.

1 required field is empty and will stay as a {{placeholder}} in the copied text.

How to use it

  1. Run this at the end of a conversation that went well, in the same chat, so the assistant can see every correction you made along the way.
  2. Tell it what will change next time. If you leave that out, it may bake this conversation's details into the template.
  3. Test the new prompt in a fresh chat with different inputs. If the result is worse, add the missing correction to the rules and test again.
  4. Save the final version where you can reach it quickly, ideally with a typed shortcut, so you actually reuse it.

Example output

Illustrative only. Real answers depend on your input and the assistant you use.

Turning points:
- You said "no jargon, and explain it to a new hire" after my first draft, which changed the tone completely.
- You asked for "three bullet points per section, maximum" after the second draft.

Reusable prompt:

You are helping me write an onboarding guide. If any essential information is missing, ask me before you write.

Topic: {{topic}}
Source material: {{source material}}
Reader: {{reader=a new hire with no background}}

Write the guide with these sections: What it is, Why it matters, How to do it, Common mistakes. Use at most three bullet points per section.

Rules: no jargon; define any term you must use; do not invent steps that are not in the source material.

Assumptions: the source material is accurate and complete; the reader will read it once.

Example of a good input: a one-page process document pasted in full.

Tips

The corrections you made are the valuable part of a conversation. Most people only keep the final answer and lose the instructions that produced it.

A template built from one conversation is a first draft. Improve it after two or three uses, when you know which parts you keep editing.

If your assistant supports it, use the same chat to ask for a shorter version once the template works, for cases where the context window is limited.

If you use HistChat, you can save the finished prompt with a typed shortcut, and any variables in double curly braces become fill-in fields when you insert it.

  • Working with AI Conversations

    Summarize a Chat for Handoff

    Asks the assistant to write a handoff brief with goal, decisions, constraints and next steps, ready to paste into a new chat.

    Customizable fields
    1 field
  • Working with AI Conversations

    Extract Decisions and Open Questions

    Pulls decisions, open questions and unstated assumptions out of a long conversation instead of summarizing all of it.

    Customizable fields
    2 fields
  • Working with AI Conversations

    Write Project Instructions for AI

    Writes a standing instruction block for a long-running AI project, with context, rules and output format you paste once and reuse.

    Customizable fields
    9 fields