Do I need a plan before starting this skill?

Yes, this skill is designed to execute an existing implementation plan. It works best when paired with a planning skill like 'superpowers:writing-plans'.

How does the two-stage review process work?

Each task is first reviewed by a spec-reviewer subagent to ensure all requirements are met. Once approved, a code-quality reviewer performs a second audit for best practices and performance.

What is the benefit of using a fresh subagent for every task?

Using a fresh subagent prevents 'context pollution,' ensuring the AI doesn't get confused by previous tasks' code or history. This leads to higher focus and fewer errors.

How is this different from the 'executing-plans' skill?

Subagent-Driven Development happens entirely within the current session for faster iteration, whereas 'executing-plans' typically involves parallel sessions.

Subagent-Driven Development

Name: Subagent-Driven Development
Author: Krosebrook

byKrosebrook

•

Productivity & Workflow

Executes implementation plans by dispatching fresh subagents for each task with a rigorous two-stage review process.

Subagent-Driven Development is a high-efficiency workflow for Claude Code designed to execute complex implementation plans within a single session. By dispatching a dedicated subagent for every task, the skill eliminates context pollution and ensures deep focus on specific requirements. Every task follows a strict two-stage review protocol—validating specification compliance first, followed by a code quality audit—to ensure production-grade output. This systematic approach combines the speed of AI automation with the rigor of professional quality gates, making it ideal for rapid feature development and codebase maintenance within large-scale projects.

Key Features

01Mandatory two-stage review (Specification Compliance and Code Quality)

021 GitHub stars

03Task-specific subagent dispatching to prevent context drift

04Continuous session execution without the need for manual handoffs

05Automated task tracking and status updates using TodoWrite

06Integrated Test-Driven Development (TDD) via implementer subagents

Use Cases

01Enforcing strict adherence to technical specifications in professional monorepos

02Executing multi-task feature plans where components are largely independent

03Rapidly prototyping complex modules while maintaining high quality standards

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add krosebrook/source-of-truth-monorepo subagent-driven-development

For use in Claude.ai and ChatGPT

Download Skill