design-debate

To facilitate a multi-agent debate for software architecture decisions, the design-debate skill spawns three AI agent roles—architect, critic, and pragmatist—to review complex designs and synthesize a structured design document after user approval.

A multi-agent AI review uses distinct roles to evaluate collaborative design from different perspectives. Use it when facing complex architectures or conflicting viewpoints to prevent echo-chamber thinking during brainstorming sessions.

Run the design-debate skill during your brainstorming session to initiate a structured design review. It requires user approval to activate and orchestrates collaboration between AI agents to generate a holistic architectural output.

Yes, multi-agent design debate features opt-in activation, allowing users to decide when to trigger the review. This avoids unnecessary escalations and ensures the three-role process is reserved for complex architecture decisions.

The best way to avoid echo-chamber thinking in software architecture reviews is to use a structured debate with distinct perspectives. Assigning architect, critic, and pragmatist roles ensures conflicting viewpoints are thoroughly evaluated.

The output from a multi-agent collaborative design debate is a structured design document. This document synthesizes the findings and interactions from the architect, critic, and pragmatist AI agents following user approval.