subagent-driven-development

Dispatch fresh subagents to execute independent tasks with two-stage reviews.

Updated Feb 2, 2026
One-click install
npx skills add https://github.com/joshua3403/giit-template --skill subagent-driven-development-joshua3403
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: subagent-driven-development
Source: https://github.com/joshua3403/giit-template/tree/main/.agent/skills/subagent-driven-development
Command: npx skills add https://github.com/joshua3403/giit-template --skill subagent-driven-development-joshua3403

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill automates executing implementation plans by dispatching fresh subagents to handle independent tasks within a single session, ensuring clean context and fast iteration.

Core Features & Use Cases

  • Fresh subagent per task to isolate context and prevent pollution.
  • Two-stage review: spec-compliance review followed by code-quality review for each task.
  • Deterministic, plan-driven execution with traceable progress and self-review.
  • Ideal for long-running plans where tasks are largely independent and require rapid turnover.

Quick Start

Use this skill to orchestrate per-task subagents for the current plan within the active session. The controller provides full task text and context to each subagent, then runs the two-stage review after each task.

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 executing an implementation plan with independent tasks?

You can automate execution by dispatching fresh subagents for each independent task within a single session, which isolates context and enables rapid iteration. The controller provides full task text and context to every subagent.

How does a two-stage review process work for subagent task dispatch?

The two-stage review enforces quality by running a spec-compliance check followed by a code-quality review after each task completes. This structured process ensures traceable progress and self-review for every dispatched subagent.

When do I need fresh subagents for plan-driven execution?

Fresh subagents are needed when executing long-running plans with largely independent tasks that require rapid turnover. Using a fresh subagent per task prevents context pollution and maintains deterministic, traceable progress.

What is the best way to isolate context for independent tasks during rapid iteration?

The best way to isolate context is dispatching a fresh subagent per task within a single-session workflow. This prevents context pollution, maintains clean execution, and enables rapid iteration for independent tasks.

Are there limitations to using subagent-driven development for task dispatch?

Subagent-driven development is limited to single-session workflows where tasks are largely independent. It is not suited for highly dependent or tightly coupled tasks that require shared context across execution steps.