How does the two-stage review process work?

Every task completed by a subagent is first reviewed for 'Spec Compliance' to ensure requirements are met, then for 'Code Quality' to ensure the implementation is clean and maintainable.

What is Subagent-Driven Development?

It is a methodology where a controller AI dispatches independent subagents to complete specific tasks from a plan, followed by rigorous automated quality checks.

When should I use this over the 'Executing Plans' skill?

Use Subagent-Driven Development when you want to stay in your current session for faster iteration and continuous progress without the overhead of parallel session handoffs.

Why use subagents instead of a single long session?

Subagents provide fresh context for each task, preventing 'context pollution' where previous tasks interfere with current logic, ensuring higher accuracy.

Does this skill support Test-Driven Development?

Yes, implementer subagents are instructed to follow TDD patterns, ensuring that every task is verified by tests and self-reviewed before being handed off for formal review.

Subagent-Driven Development

Name: Subagent-Driven Development
Author: januarfonti

byjanuarfonti

0•

生产力与工作流

Executes implementation plans by dispatching independent subagents with a mandatory two-stage review process for every task.

Subagent-Driven Development is a specialized workflow designed to execute complex implementation plans by breaking them into discrete, independent tasks. By dispatching a fresh subagent for each task within the same session, it prevents context pollution and maintains high precision. The skill enforces a rigorous two-stage review process: first, a specification compliance check to ensure all requirements are met, followed by a code quality review to maintain architectural standards. This approach is perfect for developers who want to accelerate their development cycle while maintaining the highest levels of code integrity and test coverage through automated AI feedback loops.

主要功能

01Automated task extraction and Todo management

020 GitHub stars

03Intelligent question-and-answer loops before implementation

04Built-in Test-Driven Development (TDD) enforcement

05Mandatory two-stage review (Spec Compliance and Code Quality)

06Context-isolated task execution via fresh subagents

使用场景

01Automating the quality assurance cycle for independent codebase components

02Executing multi-step feature plans within a single Claude session

03Managing complex refactoring projects that require high-precision task isolation

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add januarfonti/opencode-template-configured subagent-driven-development

For use in Claude.ai and ChatGPT

Download Skill