Agent skill
nw-liberating-structures-facilitation
Practitioner guide for selecting and specifying Liberating Structures, World Café, and Open Space Technology based on group size and learning intent
Install this agent skill to your Project
npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/other/other/nw-liberating-structures-facilitation
SKILL.md
Liberating Structures and Large-Group Facilitation
Decision Rule: Choose Structure by Learning Intent First, Then Group Size
| Learning Intent | Time Available | Best Structure | Group Size Range |
|---|---|---|---|
| Generate ideas or questions from everyone | 12 min | 1-2-4-All | Any |
| Surface what must stop before change can happen | 35 min | TRIZ | Any |
| Peer coaching on a live challenge | 30 min | Troika Consulting | Any (triads) |
| Structured reflection after a shared experience | 45 min | What-So-What-Now-What | Any (5–7 per group) |
| Transfer practitioner knowledge to larger group | 35–70 min | User Experience Fishbowl | 15–150+ |
| Connect audience to an expert through story | 35–60 min | Celebrity Interview | Any (theater-style) |
| Define minimum rules for a purpose | 35–50 min | Min Specs | Any (4–7 per table) |
| Open a session, build connections | 20 min | Impromptu Networking | Any |
| Activate individual agency without approval | 20 min | 15% Solutions | Any (2–4 per group) |
| Prioritize bold ideas from a large crowd | 30 min | 25/10 Crowdsourcing | 20–500+ |
| Cross-pollinate knowledge across silos | 2–3 hours | World Café | 12–hundreds |
| Self-organized exploration of complex theme | 90 min – 3 days | Open Space Technology | 10–1,000+ |
| Parallel topics + whole-group harvest | 50–70 min | Gallery Walk + Parallel Workstreams | 20–100+ |
The 10 Core Structures for IT Workshops
1. 1-2-4-All — Universal Idea Generator
Time: 12 minutes. Group: Any size.
Protocol: Silent individual (1 min) → pairs (2 min) → groups of 4 (4 min) → whole group (5 min, one idea per foursome).
Discipline: Enforce the 1-minute silence before pairs speak. Without it, early talkers anchor everyone's thinking.
How to specify: Frame a single, specific question. Bad: "What do you think about testing?" Good: "What is one thing stopping your team from writing tests first?"
2. TRIZ — Surface Counterproductive Habits
Time: 35 minutes. Group: 4–7 per group.
Protocol: Three 1-2-4-All rounds:
- "What would guarantee the worst possible outcome for [main objective]?"
- "Which of those things do we currently do?"
- "What will we commit to stop doing?"
Discipline: Round 3 outputs must be "stop X" — never "start Y."
IT example (Round 1): "How would we guarantee deployments fail every Friday and trust collapses?"
3. Troika Consulting — Peer Coaching in Triads
Time: 30 minutes (three 10-minute rounds). Group: Triads.
Protocol per round: Client presents challenge (1–2 min) → clarifying questions (1–2 min) → client turns away → consultants brainstorm (4–5 min) → client turns back and names top takeaway (1–2 min). Rotate.
Non-negotiable: The client must physically turn away. This eliminates their social influence on the advice.
4. What, So What, Now What (W³) — Structured Reflection
Time: 45 minutes. Group: 5–7 per group.
| Stage | Individual | Small Group | Whole Group |
|---|---|---|---|
| WHAT? (observations only) | 1 min | 2–7 min | 2–5 min |
| SO WHAT? (patterns, hypotheses) | 1 min | 2–7 min | 2–5 min |
| NOW WHAT? (actions) | 1 min | 2–7 min | 2–5 min |
Discipline: Hold them in "So What" until at least two patterns are named. Patterns precede useful actions.
5. User Experience Fishbowl — Practitioner Knowledge Transfer
Time: 35–70 minutes. Group: Inner circle 3–7; outer circle unlimited.
Protocol: Setup (2 min) → inner circle conversation (10–25 min) → outer circle formulates questions (4 min) → Q&A (10–25 min) → W³ debrief (10–15 min).
Non-negotiable: Inner circle = direct personal experience only. No rank, no theories — stories only.
IT examples: Team who completed a microservices migration. Engineers who ran a production postmortem. Developers who adopted TDD in an existing codebase.
6. Celebrity Interview — Expert Dialogue at Scale
Time: 35–60 minutes. Group: Unlimited, theater-style.
Protocol: Interviewer frames topic (3 min) → structured interview (15–30 min) → audience generates questions via 1-2-4 on index cards (5–10 min) → interviewer asks best questions (5–10 min).
Interview arc: inspiration → challenges → what keeps you going → what do we hope happens.
Discipline: Every answer must be a concrete story, not a prepared remark.
7. Min Specs — Constraint-Based Design Thinking
Time: 35–50 minutes. Group: 4–7 per table.
Protocol: Generate must-dos and must-not-dos (6 min) → reduction test: "If violated, could we still succeed? If yes, remove it" (15 min) → repeat reduction (15 min) → cross-table consolidation (15 min).
Discipline: Elimination test is binary. If participants argue, clarify purpose and retry.
8. Impromptu Networking — Session Opener
Time: 20 minutes. Group: Any size.
Protocol: Three rounds of 4–5 minutes in rotating pairs. One challenge question + one reciprocal question per round. Example: "What challenge do you bring?" + "What do you hope to contribute?"
9. 15% Solutions — Individual Agency Activation
Time: 20 minutes. Group: 2–4 per group.
Protocol: Individual identifies actions within their own discretion, requiring no additional resources or authority (5 min) → share with group (3 min/person) → peer consultation (5–7 min/person).
Discipline: Verify each solution falls within the individual's own discretion. If it requires someone else's approval, it's not a 15% solution.
10. 25/10 Crowdsourcing — Large-Group Prioritization
Time: 30 minutes. Group: 20–500+.
Protocol: Write bold idea + first action step on index card (5 min) → cards circulate in five pass-and-score rounds, each scores 1–5 (15 min) → countdown reveal of top ten (5 min).
World Café — Cross-Pollination Protocol
Time: 2–3 hours. Group: 12 to hundreds (3–5 per table).
Session structure:
| Phase | Duration |
|---|---|
| Arrival and welcome | 15 min |
| Opening and orientation | 15 min |
| Round 1 — opening question | 20–25 min |
| Round 2 — deepening question | 20–25 min |
| Round 3 — action question | 20–25 min |
| Whole-group harvest | 20–30 min |
| Closing | 10–20 min |
Table host role: One person stays per table across all rotations, briefing new arrivals on prior rounds. This is the mechanism that creates cross-pollination — non-negotiable.
Three-question arc:
- Opening: "When have we been at our best?" (shared memory)
- Deepening: "What would it take to...?" (barriers and possibilities)
- Action: "What commitment could we make today?" (ownership)
When NOT to use: When the group needs to produce a document, reach a specific decision, or when only one correct answer exists.
Open Space Technology — Self-Organizing Format
Time: 90 minutes (mini) to 3 days. Group: 10 to 1,000+.
Hard preconditions — confirm both before committing:
- Participants have genuine passion for the central theme
- Participants are willing to take personal responsibility for what they propose
Four Principles: Whoever comes are the right people | Whatever happens is the only thing that could have | Whenever it starts is the right time | When it is over, it is over.
Law of Two Feet: If you're neither learning nor contributing, move.
| Step | Duration |
|---|---|
| Opening circle | 15–30 min |
| Marketplace (participants post topics) | 15–20 min |
| Concurrent sessions (self-organized) | 60 min – multiple days |
| Closing circle | 30–60 min |
Mini-OST (90 min): Opening circle 15 min → Marketplace 15 min → Two 20-min session slots → Closing circle 15 min. Handles 4–8 topics; 10–60 participants.
When NOT to use: Fixed curriculum, specific required conclusion, mandatory participation, or leadership already has the answer.
Gallery Walk with Parallel Workstreams
Time: 50–70 minutes. Group: 20–100+.
| Phase | Duration | Activity |
|---|---|---|
| Topic assignment | 5 min | Groups of 4–6 receive distinct topic |
| Parallel workstreams | 20–30 min | Each group produces visible artifact |
| Gallery walk | 15–20 min | All read all artifacts, add sticky comments |
| Whole-group debrief | 10–15 min | Facilitator surfaces cross-workstream patterns |
Scaling: 60+ participants requires 2–3 co-facilitators. Always assign a "reporter" role per group at 60+.
Common Workshop Strings
String A: Workshop Opener (45 min): Impromptu Networking (20) → 1-2-4-All on learning goals (12) → synthesis (13)
String B: Deep Retrospective (80 min): TRIZ (35) → W³ debrief of TRIZ output (45)
String C: Expert Knowledge Transfer (90 min): Celebrity Interview or UX Fishbowl (60) → W³ debrief (30)
String D: Team Agreement Design (70 min): 1-2-4-All to generate principles (12) → Min Specs reduction (50) → 15% Solutions (20)
String E: Large-Group Discovery Day (Full Day): Impromptu Networking (20) → Celebrity Interview (45) → World Café (2hr) → 25/10 Crowdsourcing (30) → W³ closing (45)
String F: Mini Open Space (Half Day): Impromptu Networking (20) → Mini-OST (90) → Gallery Walk (60) → 15% Solutions (20)
Failure Modes to Avoid
| Mistake | Consequence | Fix |
|---|---|---|
| Skipping silent reflection in 1-2-4-All | Dominant voices anchor all output | Enforce 1-min silence; walk around while participants write |
| TRIZ without a clear objective | Round 1 is vague; no traction | State main objective explicitly before first question |
| Wrong people in Fishbowl inner circle | Audience learns nothing real | Select by direct experience only |
| Rushing So What in W³ | Plans lack grounding | Hold So What until two patterns are named |
| OST when sponsor has a predetermined answer | Participants lose trust | Use World Café instead |
| World Café without a table host role | No cross-pollination | One person always stays per table |
| Gallery Walk without sticky-note commenting | Walk is passive | Require written comments on each chart |
Virtual/Remote Adaptation
Universal Remote Principles
- Breakout rooms replace physical sub-groups
- Miro boards replace sticky notes — use Private Mode
- Shared on-screen timer required
- Add 2-3 min buffer per protocol for breakout room management
- Reduce group sizes by 1-2 vs in-person
Protocol-Specific Adaptations
| Protocol | Virtual (Zoom + Miro) |
|---|---|
| 1-2-4-All | 1: individual Miro sticky (Private Mode ON); 2: pair breakout 3 min; 4: group breakout 5 min; All: main room synthesis |
| TRIZ | Miro frame per question; sub-groups in separate sections; synthesis in main frame |
| W³ | Miro: 3 columns; individual input (Private Mode), then group reveal |
| Impromptu Networking | Zoom: random pairs in breakout, 3 rounds of 5 min each |
| Min Specs | Miro sticky clusters: "must have" vs "must not have" in separate zones |
| UX Fishbowl | Zoom: 4-person breakout + unmute rule; outer circle stays muted |
Not Adaptable Online
- Open Space Technology (requires physical movement)
- Wise Crowds (consultation rounds lose energy remotely)
- Conversation Cafe (deep dialogue suffers from video fatigue)
4C Phase Mapping
| 4C Phase | Recommended LS Protocols |
|---|---|
| C1: Connections | 1-2-4-All (prior knowledge), Impromptu Networking |
| C2: Concepts | What-So What-Now What?, User Experience Fishbowl, Min Specs |
| C3: Concrete Practice | TRIZ, 15% Solutions, Shift & Share |
| C4: Conclusions | 15% Solutions (commitments), What-So What-Now What? (debrief) |
Bloom's Alignment
- Remember/Understand: 1-2-4-All, W³ (information sharing)
- Apply: 15% Solutions, TRIZ
- Analyze/Evaluate: User Experience Fishbowl, Wise Crowds
- Create: Open Space Technology, Liberating Structures Sprint
Recommended Agent Skills
Expand your agent's capabilities with these related and highly-rated skills.
agent-ops-spec
Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.
agent-ops-state
Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.
agent-ops-spec
Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.
agent-ops-testing
Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.
agent-ops-testing
Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.
agent-ops-state
Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.
Didn't find tool you were looking for?