Plan a Sprint Retrospective with Timed Steps, Questions and a Dot-Vote

Get a retro format matched to your team's mood and problems, with a timed agenda, warm-up, open questions, and a dot-vote that ends in at most three owned actions.

You getFormat chosen and why, timed agenda table and retro questions

Prompt

Prompt
Act as an experienced agile coach who runs sprint retrospectives for software teams.

Your job: design a retrospective for my team, including the questions and a timed facilitation plan.

Before you start, ask me these questions in one message and wait for my answers:
1. How many people will attend, and is it remote, in person or hybrid?
2. How long is the retro, and how long was the sprint?
3. What is the team's mood right now (tired, frustrated, upbeat, quiet)?
4. What went wrong or caused friction this sprint?

Rules:
- Pick one retro format that fits the mood and problems, and explain in two sentences why you chose it over one named alternative.
- Every step has a time in minutes, and the times add up exactly to the total length I gave.
- Include a warm-up under 5 minutes that does not force anyone to share personal feelings.
- Write 4-6 open questions; none can be answered yes or no, and none blames a person.
- End with a dot-vote that produces at most three actions, each with one owner and a check-in date.

Output format:
1. Format chosen and why
2. Timed agenda table: Step | Minutes | What the facilitator says or does
3. Retro questions (numbered)
4. Dot-vote instructions
5. Blank action table for the room: Action | Owner | Check-in date
6. Three facilitator tips for this team's mood

Finish with one line on what could derail this retro and how to handle it.

Example output

Example for: a remote team of 7, 60-minute retro after a two-week sprint, mood tired and frustrated, problems were late scope changes and a flaky CI pipeline.

1. Format chosen and why

Sailboat (wind, anchors, rocks, island). It names blockers as things slowing the boat rather than people's failures, which suits a tired team. I chose it over Mad/Sad/Glad because that format tends to turn into venting when frustration is already high.

2. Timed agenda

StepMinutesWhat the facilitator says or does
Warm-up4"In one word or emoji, what was the weather of this sprint for you?" Typed in chat, no explanation needed.
Set the stage3Read the Prime Directive, show the sailboat board, state the goal: max three actions.
Silent writing12Everyone adds sticky notes to all four areas. Camera optional.
Group and read8Facilitator clusters duplicates aloud; authors add one line of context.
Discuss15Walk anchors and rocks first, then wind.
Dot-vote5Vote on clusters (see below).
Define actions9Turn top clusters into actions with owners.
Close4Each person rates the retro 1-5 in chat.
Total60

3. Retro questions

  1. What pushed us forward this sprint, even a little?
  2. What slowed us down most, and when did we first notice it?
  3. How did late scope changes reach us, and where could they have been caught earlier?
  4. What would make the CI pipeline something we trust again?
  5. What risk do you see coming in the next sprint?
  6. What is one thing we should keep doing exactly as is?

4. Dot-vote instructions

  • Each person gets 3 dots; they may stack them.
  • Vote only on anchors and rocks.
  • Top three clusters move to actions; ties are broken by the product owner.

5. Action table

ActionOwnerCheck-in date

6. Facilitator tips

  • Start with wind so the team doesn't open on complaints.
  • When scope talk turns to blaming the product owner, reframe: "What process let this change arrive on day 9?"
  • Keep actions small enough to finish in one sprint; "fix CI" becomes "quarantine the 4 flakiest tests."

Risk: The CI discussion can eat the whole retro; timebox it to 7 minutes and park deeper fixes for a separate session.

Written by AI from this prompt. Your results will vary.

Tips and common edits

Why this works

  • Asking about mood and recent problems first lets the model pick a format that fits instead of defaulting to start/stop/continue.
  • Requiring minutes that add up to the real retro length makes the agenda usable as a live run sheet.
  • Capping actions at three with named owners stops retros from ending in a long wish list nobody follows up on.

Use before your next sprint retro. Have the headcount, retro length, an honest read of team mood and the sprint's main problems ready. Works in ChatGPT, Claude or Gemini.

Common edits

  • Ask for a 30-minute version for short sprints or teams with heavy meeting load.
  • Request the questions formatted as sticky-note prompts for Miro, FigJam or a whiteboard.

Prompt details

By
Prompts-GPT editors
Published

Did this prompt work?

Keep exploring

More in Productivity