Back to Skills

facilitation-patterns

lyndonkl
Updated Today
24 views
5
5
View on GitHub
Designaidesign

About

This Claude Skill provides structured facilitation techniques for running effective collaborative sessions like workshops, retrospectives, and team decision-making. It helps developers manage group dynamics, ensure equal participation, and handle challenges like conflict or groupthink. Use it when planning meetings, designing agendas, or needing proven patterns for structured group discussions.

Documentation

Facilitation Patterns

Table of Contents

Purpose

Facilitation Patterns provide structured approaches for running productive group sessions—from quick standups to multi-day workshops. This skill guides you through selecting the right format, designing agendas, managing participation dynamics, making group decisions, and handling difficult situations to achieve session objectives.

When to Use

Use this skill when:

  • Planning sessions: Design meeting agenda, choose workshop format, estimate timing
  • Running meetings: Daily standup, 1:1s, team sync, all-hands, board meeting
  • Facilitating workshops: Brainstorm, design sprint, strategy offsite, training session
  • Team decisions: Need structured decision-making (voting, consensus, delegation)
  • Retrospectives: Sprint retro, project postmortem, quarterly review
  • Stakeholder engagement: Kickoff, requirements gathering, alignment session
  • Conflict resolution: Tensions between team members, opposing viewpoints, stuck decisions
  • Participation issues: Some dominate conversation, others silent, sidebar discussions, low energy
  • Remote facilitation: Virtual meetings, hybrid sessions, asynchronous collaboration

Trigger phrases: "facilitation", "workshop design", "meeting patterns", "how to run", "session planning", "effective meetings", "group discussion", "team collaboration"

What Is It?

Facilitation Patterns are proven formats and techniques for structuring group interactions to achieve specific outcomes (decisions, ideas, alignment, learning, etc.).

Core components:

  • Format selection: Which pattern fits your goal? (divergent brainstorm, convergent decision, alignment, learning)
  • Agenda design: Time-boxed activities in logical flow (diverge → converge, build → test)
  • Participation techniques: Ensure everyone contributes (round robin, silent writing, breakouts)
  • Decision methods: How to choose (consensus, consent, vote, delegation, advisory)
  • Energy management: Pacing, breaks, energizers, transitions
  • Difficult dynamics: Handle dominators, silent participants, conflict, tangents, groupthink

Quick example:

Scenario: Product team needs to prioritize features for Q2 (8 people, 90 minutes).

Pattern selected: Effort-Impact Workshop (diverge → assess → converge)

Agenda:

  1. Frame (10 min): Present context, Q2 goals, constraints
  2. Diverge (20 min): Silent brainstorm on sticky notes (what features could we build?)
  3. Cluster (15 min): Group similar ideas, clarify duplicates
  4. Assess (25 min): Plot features on effort-impact 2×2 matrix (in pairs, then discuss)
  5. Converge (15 min): Dot voting on "quick wins" (high impact, low effort)
  6. Decide (10 min): Top 5 features by votes → facilitator makes final call with input
  7. Close (5 min): Summarize decisions, next steps, feedback on session

Outcome: Prioritized list of 5 features, buy-in from team (they contributed), completed in 90 min.

Core benefits:

  • Structured productivity: Clear process prevents aimless discussion
  • Inclusive participation: Techniques ensure quiet voices heard, dominators managed
  • Better decisions: Multiple perspectives, bias mitigation (group vs individual), explicit criteria
  • Time efficiency: Time-boxing prevents overruns, focused activities
  • Psychological safety: Ground rules, equal participation, separated idea generation from evaluation

Workflow

Copy this checklist and track your progress:

Facilitation Planning Progress:
- [ ] Step 1: Define session objectives
- [ ] Step 2: Select facilitation pattern
- [ ] Step 3: Design agenda
- [ ] Step 4: Prepare materials and logistics
- [ ] Step 5: Facilitate the session
- [ ] Step 6: Close and follow up

Step 1: Define session objectives

What outcome do you need? (Decision, ideas, alignment, learning, relationship-building). Who attends? How much time? See resources/template.md.

Step 2: Select facilitation pattern

Based on objective and group size, choose pattern (Brainstorm, Decision Workshop, Alignment Session, Retro, Design Sprint). See Common Patterns and resources/methodology.md.

Step 3: Design agenda

Create time-boxed agenda with activities, transitions, breaks. Follow diverge-converge flow. See resources/template.md and resources/methodology.md.

Step 4: Prepare materials and logistics

Set up space (physical or virtual), prepare slides/boards, send pre-work if needed, test tech. See resources/template.md.

Step 5: Facilitate the session

Run agenda, manage time, ensure participation, handle dynamics, track outputs. See resources/methodology.md and resources/methodology.md.

Step 6: Close and follow up

Summarize outcomes, clarify next steps and owners, gather feedback, share notes. See resources/template.md.

Validate using resources/evaluators/rubric_facilitation_patterns.json. Minimum standard: Average score ≥ 3.5.

Common Patterns

Pattern 1: Divergent Brainstorm (Generate Ideas)

  • Goal: Maximum idea generation, creative exploration
  • Format: Silent individual brainstorm → share → cluster → refine
  • Techniques: Crazy 8s, SCAMPER, "Yes, and...", defer judgment
  • Time: 30-60 min for 5-10 people
  • Output: 30-100 ideas, clustered by theme
  • When: Need creative options, early in project, no single right answer

Pattern 2: Convergent Decision Workshop (Choose Direction)

  • Goal: Narrow options, make decision with group input
  • Format: Present options → assess criteria → vote/rank → decide
  • Techniques: 2×2 matrix (effort-impact), dot voting, affinity grouping, forced ranking
  • Time: 60-90 min for decision, 2-3 hours for complex
  • Output: Prioritized list or single decision with rationale
  • When: Multiple options exist, need buy-in, criteria clear

Pattern 3: Alignment Session (Build Shared Understanding)

  • Goal: Get everyone on same page (vision, strategy, plan)
  • Format: Present → Q&A → small group discussion → report back → synthesize
  • Techniques: Fishbowl, gallery walk, 1-2-4-All, consensus check
  • Time: 90-120 min for alignment, half-day for strategy
  • Output: Shared mental model, documented assumptions, commitments
  • When: Starting project, misalignment detected, new team formation

Pattern 4: Retrospective (Reflect and Improve)

  • Goal: Learn from experience, identify improvements
  • Format: Set context → gather data → generate insights → decide actions → close
  • Techniques: Start-Stop-Continue, Mad-Sad-Glad, Timeline, Sailboat, 4Ls (Liked, Learned, Lacked, Longed for)
  • Time: 45-90 min for sprint retro, 2-3 hours for project postmortem
  • Output: 2-5 actionable improvements with owners
  • When: End of sprint/project, recurring team practice, after incident

Pattern 5: Design Sprint (Prototype and Test)

  • Goal: Rapidly prototype and validate concept
  • Format: 5 days: Understand → Diverge → Decide → Prototype → Test
  • Techniques: Sketching, storyboarding, Crazy 8s, Heat Map voting, user testing
  • Time: 5 full days (can compress to 2-3 days)
  • Output: Validated prototype, user feedback, go/no-go decision
  • When: Big design decision, high uncertainty, time to test before committing

Pattern 6: Asynchronous Collaboration (Remote/Distributed)

  • Goal: Collaborate across time zones, allow reflection time
  • Format: Post prompt → async responses (24-48h) → sync synthesis session → document
  • Techniques: Shared docs, threaded discussions, Loom videos, async voting (Polly, Simple Poll)
  • Time: 2-5 days total (30-60 min sync session)
  • Output: Documented decisions, rationale, action items
  • When: Global team, deep thinking needed, no urgency for immediate decision

Guardrails

Critical requirements:

  1. Objectives before format: Start with "what outcome do we need?" not "let's do a brainstorm". Format serves objective. If objective unclear, session will drift.

  2. Time-box ruthlessly: Parkinson's Law (work expands to fill time). Set strict timers, end activities even if incomplete. 25 min generates better focus than open-ended "until we're done."

  3. Separate divergence from convergence: Don't critique ideas during brainstorm (kills creativity). Defer judgment. Generate first, evaluate second. Premature convergence yields safe, obvious ideas.

  4. Ensure psychological safety: Ground rules (no interrupting, critique ideas not people, Vegas rule if needed). Address power dynamics (boss speaks last, use anonymous input). Without safety, you get groupthink or silence.

  5. Manage participation actively: Silent people have ideas (use individual writing, round robin, small groups). Verbose people dominate (time limits, "let's hear from others", parking lot for tangents). Don't let dysfunction fester.

  6. Decide how decisions are made: Consensus (everyone agrees), consent (no objections), majority vote, delegation (input → decision-maker). Announce method upfront. Lack of clarity → "I thought we decided, but nothing happened."

  7. Track outputs visibly: Shared board, live doc, sticky notes. Everyone sees same thing (reduces misunderstanding). Assign scribe if needed. Invisible outputs are easily lost.

  8. Close with clarity: What was decided? Who does what by when? What's still open? How will we communicate? 5 min close prevents week of confusion.

Common pitfalls:

  • No agenda: Meetings drift, go long, participants unclear on purpose. Always have agenda (even 3 bullets).
  • Wrong people: Decision-makers absent, too many observers, missing key stakeholders. Right people > right process.
  • Too much content: 10 topics in 60 min = shallow on all. Better: 2-3 topics, go deep, make progress.
  • Facilitator dominates: Facilitator should guide process, not content. Reduce own talking, ask questions, stay neutral.
  • No breaks: 2+ hours without break → diminishing returns. Break every 60-90 min (5-10 min).
  • Ignoring energy: Pushing through low energy → poor output. Use energizers, adjust pace, or stop early.

Quick Reference

Key resources:

Decision-making methods:

  • Consensus: Everyone must agree (slow, high buy-in, use for high-stakes or high-impact decisions)
  • Consent: No one objects / "safe to try" (faster than consensus, Sociocracy)
  • Majority vote: >50% wins (quick, can leave minority feeling unheard)
  • Advisory: Input from group, decision by one person (fast, accountable, use when decision-maker clear)
  • Delegation: Empower subset to decide with constraints (scales well, trust required)

Participation techniques:

  • Round robin: Each person speaks in turn (ensures equal airtime)
  • 1-2-4-All: Think alone → pairs → fours → whole group (builds ideas, safe for introverts)
  • Silent writing: Sticky notes or shared doc, no talking (prevents groupthink, good for brainstorms)
  • Breakout rooms: Small groups (3-5 people) discuss, report back (scalable, increases participation)
  • Dot voting: Each person gets N dots to vote on ideas (quick prioritization, visual)
  • Fist to Five: Show fingers 0-5 to gauge agreement (quick temperature check)

Energizers (5-10 min):

  • Standup stretch: Literally stand and stretch (blood flow)
  • Quick icebreaker: "One word to describe how you're feeling", "What's on your desk right now?"
  • Music break: Play upbeat song, encouraged to dance/move
  • Pair share: 2 min with partner on non-work topic
  • Voting game: Thumbs up/down rapid-fire questions ("Coffee or tea?")

Timing guidelines:

  • Daily standup: 15 min (5-10 people, 1 min each)
  • 1:1: 30-60 min (half listening, half topics)
  • Team sync: 60 min (updates, 1-2 discussion topics)
  • Brainstorm: 30-60 min (diverge, cluster, dot vote)
  • Decision workshop: 90-120 min (options, criteria, discussion, vote)
  • Retrospective: 60-90 min (sprint), 2-3 hours (project)
  • Alignment session: 2-4 hours (include breaks)
  • Design sprint: 5 full days (or compressed to 2-3 days)

Red flags (adjust or stop session):

  • 50% on laptops/phones (not engaged) → take break, energizer, or change format

  • Same 2-3 people talking entire time → round robin, small groups
  • Sidebar conversations → address directly ("Let's have one conversation"), or acknowledge and parking lot
  • Confusion about purpose → stop, re-clarify objective, adjust agenda
  • Running 30+ min over → apologize, reschedule rest, or ruthlessly cut content

Inputs required:

  • Objective: What outcome do you need? (Decision, ideas, alignment, learning)
  • Participants: Who? How many? Roles? Power dynamics?
  • Time: How long? (Realistic estimate, not wishful thinking)
  • Constraints: Location (remote/in-person), budget, cultural norms

Outputs produced:

  • facilitation-plan.md: Session design (objective, agenda, materials, decision method, outputs)
  • session-notes.md: What was discussed, decisions made, action items with owners

Quick Install

/plugin add https://github.com/lyndonkl/claude/tree/main/facilitation-patterns

Copy and paste this command in Claude Code to install this skill

GitHub 仓库

lyndonkl/claude
Path: skills/facilitation-patterns

Related Skills

sglang

Meta

SGLang is a high-performance LLM serving framework that specializes in fast, structured generation for JSON, regex, and agentic workflows using its RadixAttention prefix caching. It delivers significantly faster inference, especially for tasks with repeated prefixes, making it ideal for complex, structured outputs and multi-turn conversations. Choose SGLang over alternatives like vLLM when you need constrained decoding or are building applications with extensive prefix sharing.

View skill

evaluating-llms-harness

Testing

This Claude Skill runs the lm-evaluation-harness to benchmark LLMs across 60+ standardized academic tasks like MMLU and GSM8K. It's designed for developers to compare model quality, track training progress, or report academic results. The tool supports various backends including HuggingFace and vLLM models.

View skill

llamaguard

Other

LlamaGuard is Meta's 7-8B parameter model for moderating LLM inputs and outputs across six safety categories like violence and hate speech. It offers 94-95% accuracy and can be deployed using vLLM, Hugging Face, or Amazon SageMaker. Use this skill to easily integrate content filtering and safety guardrails into your AI applications.

View skill

langchain

Meta

LangChain is a framework for building LLM applications using agents, chains, and RAG pipelines. It supports multiple LLM providers, offers 500+ integrations, and includes features like tool calling and memory management. Use it for rapid prototyping and deploying production systems like chatbots, autonomous agents, and question-answering services.

View skill