# Decision log prompt: free ChatGPT template

> Pull every settled decision out of meeting notes and messages into a dated log: who decided, why, and when to revisit, plus what still needs confirming.

Source: https://learn.ynteractive.com/content/prompts/decision-log-prompt · Updated 2026-10-01 · Free from Stepthrough (https://learn.ynteractive.com)

[Meetings and notes](https://learn.ynteractive.com/content/prompts#job-meetings-and-notes) · Works in ChatGPT, Claude, Gemini and Copilot

Updated October 1, 2026 · By [Robert Breen](https://learn.ynteractive.com/content/about)

Use this once a month, or whenever someone asks "wait, when did we decide that?" It reads notes from several meetings, texts or emails and keeps only what was actually settled, with the date, the person who decided and the reason. Things that only sound decided go on a separate list for you to confirm.

A decision log is different from an action items list. Action items are who does what by when. Decisions are the rules and choices that stay in force afterward, like how you price a job or what every bid must say.

## 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 the recordkeeper for [BUSINESS], keeping a running log of business decisions.
Context: Below are notes from [SOURCES]. A decision is something the owner or the group settled, not a suggestion or an idea still being discussed. Only log what the notes show was settled. Use the dates and names in the notes. If the reason or the person who decided isn't written down, put "Not recorded".
Task: Write one log entry for each decision: date, decision, decided by, the reason given, what it changes or replaces, and when to revisit it if anyone said. Then list anything that sounds decided but isn't, so I can confirm it.
Format: A table with the columns Date, Decision, Decided by, Reason, Replaces, Revisit. One decision per row, written as a plain statement under 20 words. Then a short list titled "Needs confirming". Ready to paste into a spreadsheet.

Here are the notes, oldest first:
[NOTES]
```

## Fill in the blanks

**`[BUSINESS]`**

Your business name and what you do.

**`[SOURCES]`**

What you are pasting: "two crew meetings and a text from the owner," "the last month of manager emails."

**`[NOTES]`**

The notes themselves, each with its date at the top. Dates are what make a log useful later.

## Example, filled in

**A made-up example** from the free lesson [ChatGPT for Electricians: Build an Electrical Bid Writer](https://learn.ynteractive.com/content/electrical-bid-writer-custom-gpt). Oak & Wire Electric is the made-up contractor in the electrical bid writer lesson, a crew of 5. The owner (we've called him Ray here) wants one place that records the rules the crew agreed on this fall.

Copy example

```text
Role: You are the recordkeeper for Oak & Wire Electric, a small residential and light-commercial electrical contractor, keeping a running log of business decisions.
Context: Below are notes from two crew meetings and a text from the owner. A decision is something the owner or the group settled, not a suggestion or an idea still being discussed. Only log what the notes show was settled. Use the dates and names in the notes. If the reason or the person who decided isn't written down, put "Not recorded".
Task: Write one log entry for each decision: date, decision, decided by, the reason given, what it changes or replaces, and when to revisit it if anyone said. Then list anything that sounds decided but isn't, so I can confirm it.
Format: A table with the columns Date, Decision, Decided by, Reason, Replaces, Revisit. One decision per row, written as a plain statement under 20 words. Then a short list titled "Needs confirming". Ready to paste into a spreadsheet.

Here are the notes, oldest first:
Sep 8 crew meeting: Ray: no more single fixed prices on bids before a site visit, use the ranges in the services guide. Two panel jobs ran over last month. Look at it again in January. Talked about a second van, Ray wants to see Q4 first.
Sep 22 crew meeting: Mike suggested we pull the permit before ordering an EV charger, Ray said OK, starting now. Maybe add Saturday emergency calls? Kim will ask the crew who'd want it.
Oct 1 text from Ray: Confirmed, every bid says who pulls the permit and that the work is inspected, starting today.
```

## What a good answer looks like

- Three rows: price ranges instead of fixed prices before a site visit (Sep 8, Ray, revisit in January), permit before ordering an EV charger (Sep 22, Ray on Mike's suggestion), and the permit and inspection line on every bid (Oct 1, Ray).
- The reason column says "Two panel jobs ran over last month" for the first row and "Not recorded" for the others, because nobody wrote one down.
- The second van and Saturday emergency calls under "Needs confirming," not in the log.
- Short statements someone can read in a year without the context.

## 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.

- For each row, find the line in your notes that shows it was settled. "Maybe" and "Ray wants to see" are not decisions.
- Check every date against the note it came from. A log with the wrong date is worse than no log.
- Look for reasons the model supplied on its own, like "to improve customer satisfaction." If you didn't write it, replace it with "Not recorded."
- Settle each "Needs confirming" item with the person who owns it before you add it.

## Why this prompt works

The Context draws a hard line between decided and discussed, and gives the model a phrase to use when a reason is missing. That keeps the log [grounded](https://learn.ynteractive.com/content/glossary/grounding) in what was really said, which is the whole value of a record you will rely on later.

## Practice it in the free lesson

**[Summarize a Meeting Transcript with ChatGPT](https://learn.ynteractive.com/content/chatgpt-summarize-meeting-transcript)**: Paste a long meeting transcript into ChatGPT and get the decisions, an action items table with owners and due dates, and a follow-up email ready to send. 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-meeting-summary)

Also useful: [ChatGPT for Electricians: Build an Electrical Bid Writer](https://learn.ynteractive.com/content/electrical-bid-writer-custom-gpt)

## Terms used here

[Structured output](https://learn.ynteractive.com/content/glossary/structured-output) · [Grounding](https://learn.ynteractive.com/content/glossary/grounding) · [Action items](https://learn.ynteractive.com/content/glossary/action-items)

## Related prompts

- [Action items table prompt](https://learn.ynteractive.com/content/prompts/action-items-table-prompt): Paste a meeting transcript or rough notes and get a table of every task, owner and due date, with "Not stated" wherever nobody…
- [Meeting recap email prompt](https://learn.ynteractive.com/content/prompts/meeting-recap-email-prompt): A copy-ready prompt that turns your checked meeting notes into a short recap email: decisions, owners and due dates, and nothing…
- [SOP writer prompt](https://learn.ynteractive.com/content/prompts/sop-writer-prompt): Turn a manager's rough notes into a numbered SOP with real roles, real numbers, safety steps first, a Check line, and CONFIRM…
- [Weekly team report prompt](https://learn.ynteractive.com/content/prompts/weekly-team-report-prompt): Turn this week's numbers and notes into a short owner report, with both weeks side by side so you can check the math, and no…

[All free prompt templates](https://learn.ynteractive.com/content/prompts) · [AI glossary](https://learn.ynteractive.com/content/glossary)
