subagent-driven-development

Execute implementation plans by dispatching subagents and running sequential spec-compliance and code-quality reviews.

9|2|Updated Jan 22, 2026
One-click install
npx skills add https://github.com/DmitriyPolukhin/antigravity-superpower --skill subagent-driven-development-dmitriypolukhin
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: subagent-driven-development
Source: https://github.com/DmitriyPolukhin/antigravity-superpower/tree/main/modules/subagent-driven-development
Command: npx skills add https://github.com/DmitriyPolukhin/antigravity-superpower --skill subagent-driven-development-dmitriypolukhin

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps you execute implementation plans by breaking work into independent chunks, running each chunk with a fresh subagent, and enforcing quality through review gates.

Core Features & Use Cases

  • Fresh subagent per task: Dispatches an implementer subagent for each task to avoid context pollution and enable parallel-safe iteration in the same session.
  • Two-stage automated review: Runs a spec compliance review first (no assumptions, code-inspection required), then a code quality review to catch maintainability/test issues.
  • Repeat-until-approved loop: If either review finds issues, the implementer fixes them and both reviews run again until approved.
  • Use Case: You have a detailed implementation plan for a feature with multiple mostly-independent tasks (e.g., adding hooks, updating modules, writing tests); you want consistent spec adherence and clean code without constant manual supervision.

Quick Start

Tell Antigravity to execute an implementation plan using Subagent-Driven Development, and it will dispatch an implementer and then run spec and code-quality reviews for 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 implementation plan execution for multi-task features?

Automated implementation plan execution dispatches a fresh subagent per task to avoid context pollution, then runs sequential spec-compliance and code-quality reviews in a fix-and-recheck loop until each task is approved.

What is subagent-driven development and how does it work?

Subagent-driven development is a workflow orchestration technique that breaks implementation plans into independent chunks, dispatching fresh implementer subagents per task and validating results through sequential spec-compliance and code-quality review gates.

How do I enforce spec compliance and code quality without manual supervision?

Enforce spec compliance and code quality through a two-stage automated review: first verifying spec adherence via code inspection without assumptions, then checking maintainability, running iterative fix-and-recheck loops until both reviews approve.

When should I use subagent workflow orchestration for implementation tasks?

Use subagent workflow orchestration when you have a detailed implementation plan with mostly-independent tasks that can be completed in the current session without context switching, requiring consistent spec adherence and clean code without constant manual supervision.

Can I use subagent-driven development for dependent or interrelated tasks?

Subagent-driven development fits best for multi-task features where tasks are mostly independent, as it dispatches fresh subagents per task and validates each result separately, making it less suitable for heavily interdependent task structures.

How does the review loop work when code quality or spec compliance issues are found?

The review loop sends issues back to the implementer subagent for fixes, then both spec-compliance and code-quality reviews run again, repeating this iterative fix-and-recheck cycle until each task is fully approved.