subagent-driven-development

Dispatch fresh subagents for isolated implementation tasks with spec and quality reviews.

2|Updated Apr 24, 2026

Search Tags:#workflow #code review #implementation #subagents #development planning #task execution

Ready to run

npx skills add https://github.com/LisztVan/subcli-cpp --skill subagent-driven-development-lisztvan

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: subagent-driven-development
Source: https://github.com/LisztVan/subcli-cpp/tree/main/.pi/skills/superpowers/skills/subagent-driven-development
Command: npx skills add https://github.com/LisztVan/subcli-cpp --skill subagent-driven-development-lisztvan

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps you execute implementation plans without losing context by splitting work into isolated tasks handled by fresh subagents.

Core Features & Use Cases

Independent Task Execution: Dispatch one subagent per task so work stays focused and isolated.
Two-Stage Review Loop: Verify each task first for spec compliance, then for code quality before marking it complete.
Continuous Plan Execution: Useful for multi-step development work where tasks can be completed in sequence without pausing for human check-ins.
Use Case: When you have a feature plan with several unrelated changes, this Skill lets you implement, review, and finish each task cleanly before moving to the next one.

Quick Start

Use this skill to break the current implementation plan into tasks, dispatch a fresh subagent for each task, and run spec and quality reviews after every 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 execute multi-step development plans without losing context?▼

Break your implementation plan into isolated tasks and dispatch a fresh subagent for each one. This approach keeps work focused and isolated, ensuring continuous plan execution without pausing for human check-ins.

How does subagent-driven development work for sequential code reviews?▼

Subagent-driven development works by running a two-stage review loop after each task: first verifying spec compliance, then checking code quality. Each completed task also requires a self-review and task tracking before moving forward.

What is the best way to implement unrelated feature changes in a single session?▼

The best way to implement unrelated feature changes is dispatching independent subagents per task. This allows you to implement, review, and finish each isolated step cleanly before moving to the next one without losing context.

Do I need a spec compliance check before code quality review?▼

Yes, a spec compliance check is required before the code quality review. This two-stage review loop ensures each task meets the original implementation plan specifications before assessing the overall code quality.

Can I use subagents for continuous plan execution without human check-ins?▼

Yes, you can use subagents for continuous plan execution without human check-ins. This Skill is designed for multi-step development work where tasks can be completed in sequence autonomously.

When should I not use isolated subagents for implementation tasks?▼

You should not use isolated subagents when your development plan involves tightly coupled steps that cannot be divided into isolated tasks. This approach requires work that can be separated into independent implementation tasks, questions, and sequential reviews.