What specialized agents are involved in the design review?

The skill spawns five specialized agents: a Product Manager for use case validation, an Architect for technical soundness, a Designer for UX/API quality, a Security Agent for threat modeling, and a CTO for codebase alignment.

How does the Design Review Gate trigger automatically?

It activates when a design document is committed to the docs/plans/ directory, immediately following the completion of a brainstorming task, or via a manual project command.

Does this skill help with task management after approval?

Yes, once all five agents approve the design, the skill can automatically create a BEADS epic and generate sequential implementation tasks for infrastructure, backend, and frontend phases.

What happens if one of the agents rejects the design document?

If any agent returns a 'NEEDS_REVISION' verdict, the skill consolidates the specific blockers and suggestions, presents them to the user for iteration, and allows for up to three re-runs before escalating to a human.

Design Review Gate

Name: Design Review Gate
Author: dsifry

bydsifry

•

コラボレーションツール

Automates multi-perspective technical design reviews by spawning specialized AI agents to validate architecture, security, and product alignment.

The Design Review Gate skill introduces a deterministic quality control layer for AI-driven software development by orchestrating a parallel review process involving five distinct specialist personas. When a design document is created or brainstorming completes, the skill automatically spawns Product Manager, Architect, Designer, Security, and CTO agents to evaluate the plan against specific rubrics. This multi-agent consensus model ensures that technical designs are robust, secure, and aligned with business goals before implementation begins, effectively catching structural flaws and security vulnerabilities early in the development lifecycle.

主な機能

01Deterministic approval logic requiring consensus from all five specialist agents

02Integrated security threat modeling and TDD readiness assessment

03Parallel multi-agent review orchestration (PM, Architect, Designer, Security, CTO)

04Automatic activation upon design document commits or brainstorming completion

0529 GitHub stars

06Automated implementation task generation via BEADS integration

ユースケース

01Validating complex technical architectures before starting large feature builds

02Standardizing design review workflows across distributed AI-assisted development teams

03Ensuring security compliance and threat modeling are addressed during the design phase

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add dsifry/goodtogo design-review-gate

For use in Claude.ai and ChatGPT

Download Skill