Why is the UX review stage separate from the code review?

Technical code reviews often focus on performance and syntax, while UX reviews focus on accessibility, feedback states, and user flow. Separating them ensures neither aspect is neglected.

What happens if a reviewer finds an issue in the code?

The workflow triggers an automated fix loop. The implementer subagent receives the feedback, applies the necessary changes, and the same reviewer re-evaluates the code until it passes the quality gate.

What is the benefit of using subagents for development?

Using a fresh subagent for each task prevents 'context pollution' where the AI becomes confused by previous unrelated code changes, leading to cleaner implementation and fewer hallucinations.

Does this skill work for web development as well as iOS?

Yes. While it includes specialized logic for iOS (Swift/SwiftUI), it also features a backend-reviewer specifically tuned for TypeScript, Supabase, and Edge Functions.

Subagent-Driven Development

Name: Subagent-Driven Development
Author: bgrober

bybgrober

•

生产力与工作流

Executes implementation plans through an automated workflow that dispatches specialized subagents for task implementation and three-stage quality reviews.

Subagent-Driven Development is a specialized workflow for Claude Code designed to handle complex implementation plans with high precision. By dispatching a fresh subagent for every individual task, the skill prevents context drift and accumulated errors. Every task undergoes a mandatory three-stage quality gate: a spec compliance review to ensure requirements are met, a technical code quality review (specialized for iOS or Backend), and a dedicated UX review to ensure accessibility and flow. This systematic approach allows developers to execute multi-task features with the confidence that every contribution is vetted for both technical excellence and user success.

主要功能

01Seamless integration with implementation plans and todo-tracking

02Isolated subagent dispatching to maintain clean context per task

03Three-stage review process: Spec Compliance, Code Quality, and UX

04Automated 'fix-and-re-review' loops for addressing reviewer feedback

051 GitHub stars

06Domain-specific reviewers for Swift/iOS and Supabase/Backend stacks

使用场景

01Executing complex feature plans with multiple independent tasks automatically

02Ensuring production-grade code quality across full-stack Swift and TypeScript environments

03Maintaining high UI/UX standards in indie projects without a dedicated design team

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add bgrober/indie-stack subagent-driven-development

For use in Claude.ai and ChatGPT

主要功能

01Seamless integration with implementation plans and todo-tracking

02Isolated subagent dispatching to maintain clean context per task

03Three-stage review process: Spec Compliance, Code Quality, and UX

04Automated 'fix-and-re-review' loops for addressing reviewer feedback

051 GitHub stars

06Domain-specific reviewers for Swift/iOS and Supabase/Backend stacks

使用场景

01Executing complex feature plans with multiple independent tasks automatically

02Ensuring production-grade code quality across full-stack Swift and TypeScript environments

03Maintaining high UI/UX standards in indie projects without a dedicated design team