Can I bypass the review if the agents keep flagging issues?

Yes. After three failed iteration cycles, the skill initiates an escalation protocol that allows a human user to manually override, defer, or cancel the feature.

How does the Design Review Gate skill activate?

The skill triggers automatically when a design document is committed to the 'docs/plans/' directory, when the brainstorming skill completes, or via the manual '/review-design' command.

What happens if the review agents do not approve the design?

If any agent returns a 'NEEDS_REVISION' verdict, the skill consolidates all feedback and presents it to the user. The gate remains closed until all five agents provide an 'APPROVED' verdict.

What is the difference between Task Mode and Team Mode?

Task Mode spawns fresh agent instances for every review round, while Team Mode creates persistent named teammates that retain context through revision cycles, reducing cold starts.

Design Review Gate

Name: Design Review Gate
Author: jpeggdev

byjpeggdev

0•

Security & Testing

Automates comprehensive technical and business design reviews by orchestrating specialized AI agent personas for unanimous approval.

The Design Review Gate skill acts as a rigorous quality assurance layer for software development by spawning a panel of specialist agents—Product Manager, Architect, Designer, Security, and CTO—to evaluate design documents. It ensures that every new feature is validated for use case clarity, technical architecture, UX/API consistency, security vulnerabilities, and TDD readiness before implementation begins. By automating the feedback loop and requiring cross-functional consensus, it prevents architectural drift and scope creep while providing a structured path from brainstorming to production-grade execution.

Key Features

01Support for Task and Team coordination modes to optimize for speed or context retention

02Integrated human escalation protocol for resolving persistent design blockers after three iterations

030 GitHub stars

04Parallel multi-agent orchestration involving PM, Architect, Designer, Security, and CTO roles

05Structured feedback reporting including specific blockers, suggestions, and clarifying questions

06Automatic activation triggered by design document commits or brainstorming completion

Use Cases

01Enforcing codebase alignment and TDD standards across distributed development teams

02Validating complex system architectures and security postures before writing code

03Ensuring product-market fit and use case clarity during the project planning phase

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add jpeggdev/buildwithjpegg design-review-gate

For use in Claude.ai and ChatGPT

Key Features

01Support for Task and Team coordination modes to optimize for speed or context retention

02Integrated human escalation protocol for resolving persistent design blockers after three iterations

030 GitHub stars

04Parallel multi-agent orchestration involving PM, Architect, Designer, Security, and CTO roles

05Structured feedback reporting including specific blockers, suggestions, and clarifying questions

06Automatic activation triggered by design document commits or brainstorming completion

Use Cases

01Enforcing codebase alignment and TDD standards across distributed development teams

02Validating complex system architectures and security postures before writing code

03Ensuring product-market fit and use case clarity during the project planning phase