Why use fresh subagents for every task?

Fresh subagents prevent 'context pollution' where previous code changes or debugging steps might confuse the AI, ensuring each task is approached with a clean slate and clear focus.

How does this differ from standard plan execution?

Unlike parallel session execution, this skill stays within your current session to avoid context-switching while still using fresh subagents to ensure clean, task-specific logic for every step of the plan.

Do I need an implementation plan to use this skill?

Yes, this skill is designed to ingest a structured implementation plan (often created by skills like superpowers:writing-plans) and execute it task-by-task.

What are the two stages of review?

Every task undergoes a Spec Compliance review first to ensure all requirements are met, followed by a Code Quality review to ensure the implementation is clean, performant, and maintainable.

Subagent-Driven Development

Name: Subagent-Driven Development
Author: claudiodearaujo

byclaudiodearaujo

0•

Productivity & Workflow

Executes complex implementation plans by dispatching fresh subagents per task with mandatory two-stage quality reviews.

Subagent-Driven Development is a specialized workflow for Claude Code designed to execute multi-task implementation plans within a single session. By isolating each task into a fresh subagent, this skill prevents context pollution and logic drift while maintaining high standards through a mandatory two-stage review process: first verifying specification compliance, then ensuring code quality. This approach bridges the gap between manual coding and fully autonomous parallel sessions, offering a high-velocity, TDD-friendly environment that catches architectural and logic errors early in the development cycle without requiring a session context switch.

Key Features

01Single-session execution to eliminate handoff and context-switch overhead

020 GitHub stars

03Deep integration with Test-Driven Development (TDD) practices

04Isolated subagent dispatching per task to prevent context pollution

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

06Automated task tracking and progress management via TodoWrite

Use Cases

01High-velocity development where quality gates and automated reviews are prioritized over manual checking

02Executing a multi-step feature plan while maintaining a continuous development session

03Implementing independent sub-tasks that require rigorous adherence to specific requirements

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add claudiodearaujo/sistema-de-narra-o-de-livro-front subagent-driven-development

For use in Claude.ai and ChatGPT

Download Skill

Key Features

01Single-session execution to eliminate handoff and context-switch overhead

020 GitHub stars

03Deep integration with Test-Driven Development (TDD) practices

04Isolated subagent dispatching per task to prevent context pollution

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

06Automated task tracking and progress management via TodoWrite

Use Cases

01High-velocity development where quality gates and automated reviews are prioritized over manual checking

02Executing a multi-step feature plan while maintaining a continuous development session

03Implementing independent sub-tasks that require rigorous adherence to specific requirements

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add claudiodearaujo/sistema-de-narra-o-de-livro-front subagent-driven-development

For use in Claude.ai and ChatGPT

Download Skill