What happens if a reviewer finds an issue?

If a reviewer identifies a gap, the implementer subagent is automatically re-engaged to fix the issue, followed by a mandatory re-review until the task meets all quality and spec gates.

How does this differ from the standard 'Executing Plans' skill?

Unlike standard execution which often uses parallel sessions, this skill stays within your current session for faster iteration while using fresh subagents for each task to prevent context pollution and 'hallucination' over long sessions.

Why are there two distinct review stages?

The first stage ensures the code actually meets the functional requirements (spec compliance), while the second focuses on maintainability, style, and architectural logic (code quality) to ensure the build is both functional and clean.

Does this skill support Test-Driven Development?

Yes, subagents dispatched by this skill are instructed to follow TDD patterns by default, ensuring that every task is verified by tests before it reaches the review stage.

Subagent-Driven Development

Name: Subagent-Driven Development
Author: mance-cyber

bymance-cyber

생산성 및 워크플로

Orchestrates the execution of complex development plans by dispatching specialized subagents for implementation and multi-stage review cycles.

소개

Subagent-Driven Development is a high-efficiency workflow for Claude Code that automates the execution of detailed implementation plans within a single session. By dispatching a fresh, context-isolated subagent for every task and enforcing a mandatory two-stage review process—checking first for specification compliance and then for code quality—it ensures high-fidelity output without the risk of context pollution. This skill is ideal for developers who want to maintain momentum while benefiting from the rigors of Test-Driven Development (TDD) and systematic quality gates, allowing for faster iteration without sacrificing software integrity.

주요 기능

Streamlined workflow from initial plan reading to final branch merging
0 GitHub stars
Automated task tracking and management via TodoWrite integration
Inherent support for Test-Driven Development (TDD) at the subagent level
Two-stage review cycle covering both specification compliance and code quality
Context isolation through fresh subagent dispatch for every individual task

사용 사례

Executing multi-step feature implementations from a pre-defined architectural plan
Refactoring large codebases where tasks are independent but require high precision
Accelerating development cycles while maintaining strict enterprise code quality standards

소개

주요 기능

Streamlined workflow from initial plan reading to final branch merging
0 GitHub stars
Automated task tracking and management via TodoWrite integration
Inherent support for Test-Driven Development (TDD) at the subagent level
Two-stage review cycle covering both specification compliance and code quality
Context isolation through fresh subagent dispatch for every individual task

사용 사례

Executing multi-step feature implementations from a pre-defined architectural plan
Refactoring large codebases where tasks are independent but require high precision
Accelerating development cycles while maintaining strict enterprise code quality standards