retrospectives
Skill1 starsChanged 7 months agoMITView on GitHub
Sprint and quarterly retrospective facilitation. Use when: running a sprint retro, facilitating a quarterly retrospective, choosing retro formats, tracking action items from retros, identifying patterns across retros, or improving team retrospective culture. Covers facilitation techniques, formats, action tracking, and anti-patterns.
In this repository42 other files it gives its agents.
- accessibility-testingSkill
- alternatives-analysisSkill
- anomaly-detectionSkill
- api-designSkill
- async-messagingSkill
- caching-strategiesSkill
- chartsSkill
- cicdSkill
- code-reviewSkill
- commsSkill
- critical-feedbackSkill
- data-analysisSkill
- databaseSkill
- data-migrationsSkill
- data-visualizationSkill
- dependency-managementSkill
- dockerSkill
- error-handlingSkill
- feature-flagsSkill
- git-workflowSkill
- goalsSkill
- ide-setupSkill
- large-changeSkill
- logging-metricsSkill
- markdown-docsSkill
- mcp-serverSkill
- meeting-agendasSkill
- okrsSkill
- presentationsSkill
- product-managementSkill
- prompt-writingSkill
- pythonSkill
- reclassificationSkill
- release-managementSkill
- roadmap-planningSkill
- securitySkill
- stakeholder-commsSkill
- statisticsSkill
- status-updatesSkill
- testingSkill
- ticket-writingSkill
- typescriptSkill
The file
We split this skill into the rules it gives an agent
2 rules. About 0 are common, 2 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 75
Best for: After a major release or project milestone
- Distinctive: line 108
Create a timeline of the sprint/quarter.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Retrospectives
When to Use
- Running a sprint retrospective
- Facilitating a quarterly or project-level retrospective
- Choosing a retro format for a specific team situation
- Following up on retro action items
- Recognizing patterns across multiple retros
- Improving a team's retro culture when retros feel stale or unproductive
Core Principles
- Psychological safety first — People must feel safe to be honest. No blame, no retaliation.
- Actionable outcomes — Every retro produces 1–3 concrete action items with owners and deadlines.
- Variety keeps engagement — Rotate formats to prevent retro fatigue.
- Follow through — Start every retro by reviewing last retro's action items. Nothing kills retro culture faster than ignored actions.
- Timebox — 60 minutes for a sprint retro, 90 minutes for a quarterly retro. Respect the clock.
Sprint Retrospective (60 minutes)
Standard Flow
| Time | Activity |
| 0:00–0:05 | Check-in: Review previous action items — completed? In progress? Dropped? |
| 0:05–0:10 | Set the stage: State the retro format, remind of ground rules |
| 0:10–0:30 | Gather data: Team writes and shares observations (format-dependent) |
| 0:30–0:45 | Generate insights: Group, discuss, and prioritize themes |
| 0:45–0:55 | Decide actions: Pick 1–3 action items with owners and due dates |
| 0:55–1:00 | Close: Quick round — one word for how you feel about the sprint |
Ground Rules (share at start)
- What's said in retro stays in retro
- Focus on process and systems, not individuals
- Everyone participates — no spectators
- Disagree respectfully; assume good intent
- The goal is improvement, not venting
Retro Formats
Rotate formats every 2–4 sprints to keep things fresh.
1. Start / Stop / Continue
Best for: General-purpose, new teams, quick retros
| Column | Prompt |
| Start | What should we start doing that we're not doing? |
| Stop | What should we stop doing because it's not helping? |
| Continue | What's working well that we should keep doing? |
2. 4Ls (Liked, Learned, Lacked, Longed For)
Best for: After a major release or project milestone
| Column | Prompt |
| Liked | What did you enjoy or appreciate? |
| Learned | What did you learn this sprint? |
| Lacked | What was missing or insufficient? |
| Longed For | What do you wish you had? |
3. Mad / Sad / Glad
Best for: Emotional check-in after a tough sprint, or when morale seems low
| Column | Prompt |
| Mad | What frustrated you? |
| Sad | What disappointed you? |
| Glad | What made you happy? |
4. Sailboat
Best for: Visual thinkers, identifying forces helping and hindering the team
Draw a sailboat:
- Wind (sails): What propelled us forward?
- Anchor: What held us back?
- Rocks (underwater): What risks do we see ahead?
- Island (destination): Where are we trying to go?
5. Timeline
Best for: Quarterly retros or after a long project
Create a timeline of the sprint/quarter. Team adds sticky notes for:
- Highs (green) — moments of success or energy
- Lows (red) — moments of frustration or failure
- Key events (blue) — incidents, launches, decisions
Discuss patterns in the timeline.
6. One-Word Check-In + Fishbowl
Best for: Large teams (8+), deep-dive on a specific topic
- Everyone shares one word describing the sprint
- Facilitator picks the dominant theme
- 3–4 volunteers discuss in a "fishbowl" (inner circle) while others observe
- Outer circle can rotate in
Quarterly Retrospective (90 minutes)
Purpose
A quarterly retro looks at patterns across sprints, not individual sprint issues. It's about systemic improvements.
Agenda
| Time | Activity |
| 0:00–0:10 | Review quarterly OKR scores and delivery metrics |
| 0:10–0:20 | Review action items from previous quarterly retro |
| 0:20–0:50 | Timeline exercise across the quarter — highs, lows, key events |
| 0:50–1:10 | Theme discussion — What patterns do we see? What systemic issues? |
| 1:10–1:25 | Action items — 2–4 systemic improvements with owners |
| 1:25–1:30 | Close — Appreciation round (each person appreciates one teammate) |
Action Item Tracking
Format
## Retro Action Items — Sprint [N] / Q[N]
| # | Action | Owner | Due Date | Status |
|---|--------|-------|----------|--------|
| 1 | [Action] | @name | YYYY-MM-DD | Open |
| 2 | [Action] | @name | YYYY-MM-DD | Open |
| 3 | [Action] | @name | YYYY-MM-DD | Open |Rules
- Max 3 action items per retro — More than 3 won't get done
- Every action has an owner — Shared ownership means no ownership
- Every action has a due date — "Sometime" means never
- Review at next retro — First agenda item, every time
- Track completion rate — If <50% of actions complete, you have a follow-through problem
Facilitation Tips
- Rotate the facilitator — Don't always facilitate yourself as EM. Give team members the opportunity.
- Use a timer — Visible countdown keeps discussions on track.
- Silent brainstorming first — Give 5 minutes of silent writing before discussion. This prevents anchoring and gives introverts space.
- Dot vote for prioritization — Each person gets 3 votes to place on themes. Discuss the top 2–3.
- Remote-friendly tools — Use a collaborative whiteboard tool (Miro, FigJam, or equivalent) for distributed teams.
- As the manager, speak last — Your opinions carry weight; let the team share first.
Anti-Patterns
| Anti-Pattern | Problem | Better Approach |
| Same format every time | Team tunes out, engagement drops | Rotate formats every 2–4 sprints |
| No action items | Retro becomes a venting session | Always end with 1–3 actions with owners |
| Actions never reviewed | Trust in the retro process erodes | Start every retro reviewing last retro's actions |
| Manager dominates | Team stops sharing honestly | Manager speaks last, facilitator rotates |
| Blame individuals | Kills psychological safety | Focus on process and systems, not people |
| Skipping retros | Problems compound, morale drops | Retros are non-negotiable; reschedule, don't cancel |
| Too many action items | Nothing gets done | Cap at 3; do fewer things well |
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

