Why is a fresh subagent used for every task?

Using a fresh subagent ensures a clean context window for every specific task, which significantly reduces the risk of 'hallucinations' or confusion caused by irrelevant previous code changes.

Does this skill support Test-Driven Development?

Yes, implementation subagents are specifically instructed to follow TDD principles by writing tests and verifying their work before submitting it for the review phase.

Can I use this for tightly coupled tasks?

It is recommended for tasks that are mostly independent. For tightly coupled tasks where the code must be written simultaneously, manual execution or a single subagent session may be more appropriate.

How does this differ from the standard executing-plans skill?

Unlike executing-plans which often utilizes parallel sessions, this skill operates within your current session using fresh subagents for each task to prevent context pollution while allowing for immediate human oversight.

Subagent-Driven Development

Name: Subagent-Driven Development
Author: jrc1883

byjrc1883

0•

Productivity & Workflow

Orchestrates implementation plans by dispatching fresh subagents for individual tasks with automated code review gates.

The Subagent-Driven Development skill streamlines complex software implementation within Claude Code by breaking plans into isolated tasks. It dispatches a fresh subagent for every task to prevent context pollution and memory drift, ensuring each component is built with a clean slate. By enforcing automated code reviews between every task, it maintains high quality standards and catches issues early in the development cycle, enabling rapid, production-grade iteration without the need for manual session handoffs.

Key Features

01Integrated task tracking via Todo lists for clear progress visibility

02Automated code review checkpoints between implementation steps

03Same-session execution for seamless developer interaction

04Inherent support for Test-Driven Development (TDD) workflows

05Fresh subagent context per task to eliminate memory pollution

060 GitHub stars

Use Cases

01Executing multi-step feature implementation plans with high precision

02Automating high-quality code generation with mandatory quality gates

03Performing complex refactors that require isolated testing and review

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add jrc1883/popkit-claude pop-subagent-dev

For use in Claude.ai and ChatGPT

Download Skill