Use when facing decisions with multiple legitimate perspectives and inherent tensions. Invoke when stakeholders have competing priorities (growth vs. sustainability, speed vs. quality, innovation vs. risk), need to pressure-test ideas from different angles before committing, exploring tradeoffs between incompatible values, synthesizing conflicting expert opinions into coherent strategy, or surfacing assumptions that single-viewpoint analysis would miss.
When facing decisions with competing stakeholder priorities (growth vs. sustainability, speed vs. quality), use this to pressure-test ideas from multiple angles. It surfaces hidden assumptions and synthesizes conflicting expert opinions into coherent, actionable strategies that acknowledge tradeoffs.
/plugin marketplace add lyndonkl/claude/plugin install lyndonkl-thinking-frameworks-skills@lyndonkl/claudeThis skill inherits all available tools. When active, it can use any tool Claude has access to.
resources/evaluators/rubric_chain_roleplay_debate_synthesis.jsonresources/examples/build-vs-buy-crm.mdresources/methodology.mdresources/template.mdCopy this checklist and track your progress:
Roleplay → Debate → Synthesis Progress:
- [ ] Step 1: Frame the decision and identify roles
- [ ] Step 2: Roleplay each perspective authentically
- [ ] Step 3: Structured debate between viewpoints
- [ ] Step 4: Synthesize into coherent recommendation
- [ ] Step 5: Validate synthesis quality
Step 1: Frame the decision and identify roles
State the decision clearly as a question, identify 2-5 stakeholder perspectives or roles that have legitimate but competing interests, and clarify what a successful synthesis looks like. See Decision Framing for guidance on choosing productive roles.
Step 2: Roleplay each perspective authentically
For each role, articulate their position, priorities, concerns, and evidence. Genuinely advocate for each viewpoint without strawmanning. See Roleplay Guidelines for authentic advocacy techniques and use resources/template.md for complete structure.
Step 3: Structured debate between viewpoints
Facilitate direct clash between perspectives on key points of disagreement. Surface tensions, challenge assumptions, test edge cases, and identify cruxes (what evidence would change each perspective's mind). See Debate Structure for debate formats and facilitation techniques.
Step 4: Synthesize into coherent recommendation
Integrate insights from all perspectives into a unified decision that acknowledges tradeoffs, incorporates valid concerns from each viewpoint, and explains what's being prioritized and why. See Synthesis Patterns for integration approaches and resources/template.md for synthesis framework. For complex multi-stakeholder decisions, see resources/methodology.md.
Step 5: Validate synthesis quality
Check synthesis against resources/evaluators/rubric_chain_roleplay_debate_synthesis.json to ensure all perspectives were represented authentically, debate surfaced real tensions, synthesis is coherent and actionable, and no perspective was dismissed without engagement. See When NOT to Use This Skill to confirm this approach was appropriate.
Good role selection:
Typical role patterns:
How many roles:
Strong framing:
Weak framing:
Each role should:
Avoiding strawmen:
❌ "The engineer just wants to use shiny new tech" (caricature)
✅ "The engineer values maintainability and believes new framework reduces technical debt"
❌ "Sales only cares about closing deals" (dismissive)
✅ "Sales is accountable for revenue and sees this feature as critical for competitive positioning"
Empathy without capitulation: You can deeply understand a perspective without agreeing with it. Each role should be the "hero of their own story."
For each role, answer:
Debate formats:
1. Point-Counterpoint
2. Devil's Advocate
3. Constructive Confrontation
4. Crux-Finding
1. Weighted Synthesis
2. Sequencing
3. Conditional Strategy
4. Hybrid Approach
5. Reframing
6. Elevating Constraint
Strong synthesis:
Weak synthesis:
Decision: Should we rewrite our monolith as microservices?
Roles:
Synthesis: Don't rewrite everything, but extract the 2-3 services with clear scaling needs (authentication, payment processing) as independent microservices. Keep core business logic in monolith for now. This addresses scalability concerns for bottleneck components (Scalability Engineer), limits risk and timeline (Pragmatic Engineer), and reduces cost to $400K vs. $2M (Finance). Revisit full migration if extracted services succeed and prove the pattern.
For a complete worked example with detailed roleplay, debate, and synthesis, see:
Skip roleplay-debate-synthesis when:
❌ Single clear expert: If one person has definitive expertise and others defer, just ask the expert ❌ No genuine tension: If stakeholders actually agree, debate is artificial ❌ Values cannot be negotiated: If ethical red line, don't roleplay the unethical side ❌ Time-critical decision: If decision must be made in minutes, skip full debate ❌ Implementation details: If decision is "how" not "whether" or "what", use technical collaboration not debate
Use simpler approaches when:
For complex multi-stakeholder decisions, see resources/methodology.md for:
Creating algorithmic art using p5.js with seeded randomness and interactive parameter exploration. Use this when users request creating art using code, generative art, algorithmic art, flow fields, or particle systems. Create original algorithmic art rather than copying existing artists' work to avoid copyright violations.
Applies Anthropic's official brand colors and typography to any sort of artifact that may benefit from having Anthropic's look-and-feel. Use it when brand colors or style guidelines, visual formatting, or company design standards apply.
Create beautiful visual art in .png and .pdf documents using design philosophy. You should use this skill when the user asks to create a poster, piece of art, design, or other static piece. Create original visual designs, never copying existing artists' work to avoid copyright violations.