This skill should be used when the user asks about "multi-perspective analysis", "multiple viewpoints", "council deliberation", "dialectical synthesis", "different perspectives on", "diverse viewpoints", "devil's advocate", "red team this", "stress test this idea", "poke holes in this", or wants to analyze a question from multiple cognitive angles simultaneously.
PolyClaude spawns parallel cognitive perspectives to analyze questions, plans, and ideas from multiple angles, then synthesizes findings through structured dialectical analysis.
The /polyclaude command is valuable when:
Six cognitive perspectives are available. The User Advocate is always included (unless explicitly excluded); others are selected adaptively based on the question type:
| Perspective | Lens | Always Asks |
|---|
| User Advocate (default) | Experience, empathy, accessibility | "How does this feel to encounter?" |
| Architect | Systems, structure, scalability | "What are the load-bearing assumptions?" |
| Skeptic | Risks, assumptions, failure modes | "What are we not seeing?" |
| Pragmatist | Simplicity, effort, trade-offs | "What's the simplest thing that works?" |
| Innovator | Alternatives, inversions, creativity | "What would the opposite look like?" |
| Temporal Analyst | Timelines, sequencing, second-order effects | "What does this look like in 6 months?" |
/polyclaude Should we rewrite the auth system or patch it?
/polyclaude --quick Is this API design reasonable? # 2 perspectives — fast sanity check
/polyclaude Should we use Redis or Postgres for sessions? # 4 perspectives — default
/polyclaude --full What's our mobile strategy? # 6 perspectives — comprehensive
/polyclaude --deep What architecture should we use? # Opus agents — maximum depth
/polyclaude --full --deep Should we pivot to B2B? # All 6 on Opus — thorough
/polyclaude --include temporal,innovator Should we refactor the API?
/polyclaude --exclude pragmatist What if we open-sourced this?
| Mode | Perspectives | Sonnet (default) | Opus (--deep) |
|---|---|---|---|
--quick | 2 | ~$0.15 | ~$0.75 |
| (default) | 4 | ~$0.30 | ~$1.50 |
--full | 6 | ~$0.45 | ~$2.25 |
A Council Report with:
Quick councils (2 perspectives) use a compact format: Verdict, Agreement/Disagreement, Blind Spots, Next Steps.
This is not "ask N AI agents the same question." It is structured dialectical analysis: