subagent-driven-development

Execute implementation plans by dispatching subagents with two-stage reviews.

Updated Mar 10, 2026
One-click install
npx skills add https://github.com/dotlab-hq/medisync --skill subagent-driven-development-dotlab-hq
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: subagent-driven-development
Source: https://github.com/dotlab-hq/medisync/tree/main/.agents/skills/subagent-driven-development
Command: npx skills add https://github.com/dotlab-hq/medisync --skill subagent-driven-development-dotlab-hq

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill streamlines the software development process by breaking down implementation plans into independent tasks, each handled by a dedicated AI subagent, ensuring high quality and rapid iteration through a structured review process.

Core Features & Use Cases

  • Task Decomposition: Automatically breaks down complex implementation plans into manageable, independent tasks.
  • Subagent Dispatch: Assigns each task to a fresh AI subagent for focused implementation.
  • Two-Stage Review: Implements a rigorous review process including spec compliance and code quality checks after each task.
  • Iterative Refinement: Manages review feedback loops, ensuring issues are fixed before proceeding.
  • Use Case: When you have a detailed feature plan, this Skill can autonomously execute each step, from initial coding to final review, significantly accelerating development cycles.

Quick Start

Use the subagent-driven-development skill to execute the implementation plan located at ./docs/plans/feature-plan.md.

Frequently Asked Questions about subagent-driven-development

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automate software development implementation plans with AI agents?

You can automate implementation plans by dispatching independent AI subagents per task, using a two-stage review process for spec compliance and code quality to manage iterative refinement within a single session.

What is the best way to break down complex feature plans into manageable tasks?

Task decomposition automatically breaks down complex implementation plans into independent tasks, assigning each to a dedicated AI subagent for focused implementation and rapid iteration.

How does the two-stage review process work for AI-generated code?

The two-stage review process implements rigorous checks for spec compliance and code quality after each task, managing review feedback loops to ensure issues are fixed before proceeding.

Does subagent-driven development require integration with git worktrees?

Yes, executing implementation plans requires integration with git worktrees, plan writing, code review, and branch finishing superpowers to manage task decomposition and subagent interaction.

Can I execute a feature plan located in a specific markdown file autonomously?

Yes, you can execute an implementation plan located at a specific file path like ./docs/plans/feature-plan.md, autonomously handling each step from initial coding to final review.

What are the limitations of using subagents for iterative development?

The approach manages iterative refinement within a single session and relies on breaking down implementation plans into independent tasks, which may not suit highly interdependent or stateful task structures.