How do the different agents contribute to the discussion?

The Critic identifies security risks and edge cases, the Creative focuses on UX and innovation, and the Pragmatist evaluates implementation costs and feasibility.

What is the Brainstorm skill for Claude Code?

It is a specialized plugin that orchestrates multiple AI agents—specifically a Critic, Creative, and Pragmatist—to debate an idea and provide a balanced recommendation.

Can I control how long the brainstorm lasts?

Yes, you can specify the number of rounds (1-3). One round provides quick perspectives, while three rounds allow for a deep cross-debate between the agents.

How do I trigger the brainstorming process?

You can invoke it directly using the 'brainstorm' command or naturally by asking for a debate or multi-agent discussion within Claude Code.

What output formats are supported?

The skill can output a summary of the debate, a structured implementation plan using EnterPlanMode, or a list of tasks using TodoWrite.

Multi-Agent Brainstorm

Name: Multi-Agent Brainstorm
Author: noin-ai

bynoin-ai

•

协作工具

Orchestrates a structured debate between specialized AI agents to provide critical, creative, and pragmatic perspectives on any topic.

关于

The Brainstorm skill enhances Claude Code by simulating a professional team environment where ideas are rigorously tested. It launches three distinct agents—a Critic, a Creative, and a Pragmatist—to evaluate your topic from the perspectives of risk, innovation, and feasibility. Through configurable rounds of cross-debate, the skill synthesizes conflicting viewpoints into a coherent summary, structured implementation plan, or actionable task list, ensuring your technical decisions are well-vetted and comprehensive.

主要功能

Automated synthesis of consensus and key debates into structured reports
1 GitHub stars
Multi-perspective debate featuring Critic, Creative, and Pragmatist personas
Configurable depth with support for 1-3 rounds of agent interaction
Flexible output formats including summaries, Todo lists, and PlanMode integration
Real-time progress tracking of brainstorming rounds using native tool calls

使用场景

Converting high-level project concepts into actionable developer todo lists
Ideating UX improvements while balancing design elegance with technical debt
Architectural review of new features to identify security risks and implementation hurdles

关于

主要功能

Automated synthesis of consensus and key debates into structured reports
1 GitHub stars
Multi-perspective debate featuring Critic, Creative, and Pragmatist personas
Configurable depth with support for 1-3 rounds of agent interaction
Flexible output formats including summaries, Todo lists, and PlanMode integration
Real-time progress tracking of brainstorming rounds using native tool calls

使用场景

Converting high-level project concepts into actionable developer todo lists
Ideating UX improvements while balancing design elegance with technical debt
Architectural review of new features to identify security risks and implementation hurdles