subagent-driven-development

Dispatch a fresh subagent for each task and review code between tasks.

1|1|Updated Sep 22, 2025
One-click install
npx skills add https://github.com/SkogAI/skillsandknowledge --skill subagent-driven-development-skogai
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: subagent-driven-development
Source: https://github.com/SkogAI/skillsandknowledge/tree/main/actions/subagent-driven-development
Command: npx skills add https://github.com/SkogAI/skillsandknowledge --skill subagent-driven-development-skogai

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill enables rapid progress in software tasks by dispatching a fresh subagent for each task within the current session, ensuring isolated context and built-in quality gates through code reviews between tasks.

Core Features & Use Cases

  • Fresh subagent per task to prevent context pollution and enable clear ownership.
  • Automatic code review between tasks to catch issues early and improve quality.
  • Structured, iterative workflow well-suited for independent tasks in a single session.

Quick Start

Load a plan file that lists the tasks, then dispatch a fresh subagent for each task and perform a code review after each task.

  • Load Plan: Read the plan and build a Todo list of tasks.
  • For each task:
  • Dispatch a new subagent with the task instructions.
  • After task completion, dispatch a code-review subagent with a summary and prior/current state.
  • Mark the task complete and proceed to the next one.
  • After all tasks, run the final code review.

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 task execution with isolated context for each step?

Automate task execution by dispatching a fresh subagent for each task, ensuring per-task isolation to prevent context pollution and enable clear ownership throughout the workflow.

What is the best way to run automated code reviews between independent tasks?

The best way to run automated code reviews is to dispatch a code-review subagent after each task completion, providing it with a summary and the prior and current state to catch issues early.

How do I orchestrate an agile workflow for a plan with mostly independent tasks?

Orchestrate an agile workflow by loading a plan file to build a task list, dispatching a new subagent for each task, and running a final code review after all tasks are marked complete.

When should I use a fresh subagent per task in my development session?

Use a fresh subagent per task when your plan contains mostly independent tasks and you need rapid iteration, as this approach enforces per-task isolation and built-in quality gates.

Can I run a final code review automatically after completing all plan tasks?

Yes, you can run a final code review automatically by dispatching a code-review subagent after all tasks in the loaded plan are marked complete and the iterative workflow finishes.

Why does dispatching a fresh subagent prevent context pollution during task execution?

Dispatching a fresh subagent prevents context pollution because it enforces strict per-task isolation, giving each task a clean context window and clear ownership of its execution.