Can I skip roles during the review process?

No, the skill follows a strict operating model where the Skeptic, Constraint Guardian, and User Advocate must all be invoked to ensure no blind spots remain before the Arbiter approves the design.

When should I use this skill in my workflow?

You should use this skill when a design or idea requires higher confidence, risk reduction, or a formal review before moving into the implementation phase.

What is the purpose of the Decision Log?

The Decision Log is a mandatory artifact that records every decision made, alternatives considered, and how reviewer objections were resolved to ensure transparency and future maintainability.

How does Multi-Agent Brainstorming differ from regular brainstorming?

Unlike parallel brainstorming which focuses on idea generation, this skill uses a sequential process with specific, constrained roles like the Skeptic and Constraint Guardian to provide rigorous critique and validation.

Multi-Agent Brainstorming

Name: Multi-Agent Brainstorming
Author: claudiodearaujo

byclaudiodearaujo

0•

Productivity & Workflow

Orchestrates a structured, multi-role peer review process to validate complex designs and identify potential risks.

Multi-Agent Brainstorming transforms single-agent designs into robust, production-ready solutions by simulating a formal peer-review process within Claude. By assigning specific, non-overlapping roles—such as the Skeptic, Constraint Guardian, and User Advocate—this skill surfaces hidden assumptions, identifies failure modes, and enforces non-functional constraints before implementation. It provides a gated, sequential workflow that prevents premature convergence and ensures high-stakes technical decisions are thoroughly stress-tested and documented.

Key Features

010 GitHub stars

02Automated risk identification and assumption testing via the Skeptic agent

03Mandatory Decision Log for tracking design rationale and resolved objections

04Formal arbitration phase to resolve conflicts and finalize architecture

05Sequential multi-agent review with constrained, non-overlapping roles

06Enforcement of non-functional constraints including performance, security, and scalability

Use Cases

01Validating a complex system architecture before starting implementation

02Stress-testing a technical proposal to surface hidden assumptions and edge cases

03Identifying security or performance bottlenecks in a proposed feature design

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add claudiodearaujo/sistema-de-narra-o-de-livro-front multi-agent-brainstorming

For use in Claude.ai and ChatGPT

Download Skill

Key Features

010 GitHub stars

02Automated risk identification and assumption testing via the Skeptic agent

03Mandatory Decision Log for tracking design rationale and resolved objections

04Formal arbitration phase to resolve conflicts and finalize architecture

05Sequential multi-agent review with constrained, non-overlapping roles

06Enforcement of non-functional constraints including performance, security, and scalability

Use Cases

01Validating a complex system architecture before starting implementation

02Stress-testing a technical proposal to surface hidden assumptions and edge cases

03Identifying security or performance bottlenecks in a proposed feature design

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add claudiodearaujo/sistema-de-narra-o-de-livro-front multi-agent-brainstorming

For use in Claude.ai and ChatGPT

Download Skill