How does the two-stage review process work?

Every task is first checked by a Spec Reviewer to ensure it meets requirements exactly. Once approved, it passes to a Code Reviewer who evaluates naming, structure, and performance.

What is Subagent-Driven Development?

It is a methodology where a Lead agent coordinates specialized subagents (Implementers and Reviewers) to execute, verify, and finalize software tasks in parallel or sequence.

What happens if a reviewer finds an issue?

The reviewer DMs the specific Implementer with feedback. The Implementer fixes the issue and resubmits for a secondary review until all standards are met.

Does this skill support Test-Driven Development (TDD)?

Yes, Implementers are explicitly instructed to follow TDD principles, writing and running tests before committing their work for review.

Can it run without experimental agent teams?

Yes, the skill includes a legacy fallback mode that executes tasks sequentially using individual subagents if the Agent Teams feature is not available.

Subagent-Driven Development

Name: Subagent-Driven Development
Author: GoCodeAlone

byGoCodeAlone

0•

コラボレーションツール

Automates plan execution using specialized agent teams for parallel software development and rigorous two-stage code review.

This skill transforms the software development lifecycle by orchestrating a hierarchical team of specialized AI agents to execute complex implementation plans. By delegating tasks to dedicated Implementers, Spec Reviewers, and Code Reviewers, it ensures high-quality output through parallel execution and a mandatory two-stage verification process. Whether using modern Agent Teams or sequential fallback modes, it enforces Test-Driven Development (TDD) and maintains strict adherence to design specifications, making it ideal for large-scale feature development and systematic refactoring within the Claude Code environment.

主な機能

010 GitHub stars

02Automated task management via shared lists and inter-agent communication

03Parallel task execution for faster implementation of complex feature sets

04Built-in Test-Driven Development (TDD) enforcement for all implementation tasks

05Role-based agent orchestration with Lead, Implementer, and Reviewer personas

06Two-stage review process verifying specification compliance before code quality

ユースケース

01Automated execution of multi-step architectural migration plans

02Scaling complex software feature implementation across large codebases

03Rigorous code quality enforcement for critical production systems

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add gocodealone/claude-superpowers subagent-driven-development

For use in Claude.ai and ChatGPT

主な機能

010 GitHub stars

02Automated task management via shared lists and inter-agent communication

03Parallel task execution for faster implementation of complex feature sets

04Built-in Test-Driven Development (TDD) enforcement for all implementation tasks

05Role-based agent orchestration with Lead, Implementer, and Reviewer personas

06Two-stage review process verifying specification compliance before code quality

ユースケース

01Automated execution of multi-step architectural migration plans

02Scaling complex software feature implementation across large codebases

03Rigorous code quality enforcement for critical production systems

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add gocodealone/claude-superpowers subagent-driven-development

For use in Claude.ai and ChatGPT