Retrospective Facilitation Frameworks Available in pm-skills: The Complete Guide

The pm-skills repository ships with three core retrospective facilitation frameworks—Start/Stop/Continue, 4Ls, and Sailboat—that can be invoked via the /sprint retro CLI command to automate sprint reviews and generate actionable feedback loops.

The phuryn/pm-skills repository provides structured tools for product management execution, including a dedicated retro skill for agile teams. These retrospective facilitation frameworks help teams surface insights, identify blockers, and create concrete improvement plans at the end of each sprint.

The Three Retrospective Facilitation Frameworks

The retro skill defined in pm-execution/skills/retro/SKILL.md implements three distinct methodologies. Each framework offers a unique lens for analyzing sprint performance and team dynamics.

Start/Stop/Continue

This framework organizes feedback into three columns: Start (what the team should begin doing), Stop (what should cease immediately), and Continue (what is working and should remain unchanged). It is ideal for teams needing clear, actionable behavioral shifts and a quick three-column view of priorities.

4Ls (Liked/Learned/Lacked/Longed For)

The 4Ls framework captures a balanced spectrum of team experience: Liked (enjoyable aspects), Learned (new knowledge acquired), Lacked (missing resources or support), and Longed For (desired improvements). Use this when you need to surface deeper learning and acknowledge both positive outcomes and gaps.

Sailboat

The Sailboat metaphor frames the retrospective as a journey: Wind (forces propelling the team forward), Anchor (elements holding the team back), Rocks (potential risks ahead), and Island (the goal destination). This works best when teams need to visualize forces, assess risks, and align on long-term objectives during remote or visual facilitation sessions.

How to Invoke Frameworks via CLI

The retro skill is exposed through the /sprint retro command defined in pm-execution/commands/sprint.md. You can let the tool auto-select a framework based on input context, or force a specific format using the --format flag.

Basic invocation with automatic framework selection:

/sprint retro "Here are the sprint notes and velocity chart"

Force a specific framework:


# Start/Stop/Continue format

/sprint retro --format "Start/Stop/Continue" "Team feedback attached"

# 4Ls format

/sprint retro --format "4Ls" "Collected user survey results"

# Sailboat format

/sprint retro --format "Sailboat" "Retrospective notes from last sprint"

Processing Pipeline and Output

When invoked, the retro skill analyzes input files (such as sticky-note CSVs, velocity charts, or previous retro markdown) and executes a four-step process:

  1. Groups similar items into thematic clusters.
  2. Highlights the most frequent topics and sentiment patterns.
  3. Generates an action-item table with priority, owner, deadline, and success metric.
  4. Outputs a markdown summary suitable for repository storage or team distribution.

Implementation Details

The framework definitions live in pm-execution/skills/retro/SKILL.md, which specifies the structure for each facilitation method. The CLI wiring in pm-execution/commands/sprint.md maps the /sprint retro command to this skill, while pm-execution/README.md provides module-level documentation. The root README.md contextualizes these tools within the broader pm-skills toolbox for product management workflows.

Summary

  • The pm-skills repository provides three built-in retrospective facilitation frameworks: Start/Stop/Continue, 4Ls, and Sailboat.
  • Frameworks are defined in pm-execution/skills/retro/SKILL.md and invoked via the /sprint retro CLI command.
  • Use the --format flag to force a specific framework, or allow automatic selection based on input context.
  • Each execution generates themed clusters, sentiment analysis, and actionable task tables with ownership and deadlines.

Frequently Asked Questions

How do I choose which retrospective facilitation framework to use?

Select Start/Stop/Continue when you need immediate behavioral changes and clear action items. Choose 4Ls for comprehensive feedback that balances successes with learning opportunities. Use Sailboat when discussing risks, impediments, and long-term strategic goals with visual metaphors.

Can I customize the retrospective formats beyond the three built-in options?

The current implementation in pm-execution/skills/retro/SKILL.md supports the three predefined frameworks. While the CLI accepts various input file types (CSVs, markdown, charts), the --format flag currently restricts selection to Start/Stop/Continue, 4Ls, or Sailboat.

What input data can the retro skill process?

The skill accepts sprint notes, velocity charts, sticky-note CSVs, user survey results, and previous retrospective markdown files. It analyzes these inputs to group themes, identify sentiment patterns, and generate the action-item table.

Where is the retrospective framework logic defined in the source code?

The framework definitions and selection logic reside in pm-execution/skills/retro/SKILL.md. The command-line interface mapping is located in pm-execution/commands/sprint.md, which wires the /sprint retro command to the skill implementation.

Have a question about this repo?

These articles cover the highlights, but your codebase questions are specific. Give your agent direct access to the source. Share this with your agent to get started:

Share the following with your agent to get started:
curl -s "https://instagit.com/install.md"

Works with
Claude Codex Cursor VS Code OpenClaw Any MCP Client

Maintain an open-source project? Get it listed too →