team-brainstorm_skill

Overview

This skill runs deep brainstorming using an adversarial agent team that debates, challenges, and refines an idea. It spawns independent teammates (Devil's Advocate, Optimist, Creative Explorer, Researcher, and optionally an Architect) to produce a richer, adversarial analysis than single-agent brainstorming. Expect higher token use and a structured synthesis suitable for decision-making or planning.

How this skill works

After capturing the core idea, the lead asks 2–4 Socratic clarification rounds to refine scope, assumptions, success criteria, and constraints. The lead then creates a team, assigns focused tasks to each teammate, and coordinates parallel research and cross-team debates. Once everyone submits findings, the lead synthesizes agreements, contentions, validated strengths, real risks, creative alternatives, and an evidence-backed recommendation, then writes a structured report and tears down the team.

When to use it

• You need adversarial, multi-perspective analysis for high-stakes decisions
• When premortem-style failure analysis is required
• To surface hard-to-see risks and counterarguments before committing resources
• When you want multiple independent analyses and cross-debate, not a single-agent summary
• If you have project or codebase context and need technical feasibility input

Best practices

• Be prepared to answer 2–4 focused clarification questions before the team launches
• Provide any relevant project or codebase context to enable the Architect/Researcher
• Define clear success criteria and constraints up front to keep the team aligned
• Allow the Devil's Advocate to push ideas—that tension yields stronger conclusions
• Expect and budget for higher token usage; reserve this for decisions that justify deeper analysis

Example use cases

• Evaluating a risky product pivot where business and technical trade-offs matter
• Exploring multiple implementation paths for a critical feature in an existing codebase
• Running a premortem on a proposed launch to surface likely failure modes and mitigations
• Generating creative alternatives and selecting one that balances risk, value, and feasibility
• Producing a documented synthesis to hand off to planning or architecture teams