subagent-driven-development

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

20|1|Updated Dec 10, 2025
One-click install
npx skills add https://github.com/galangryandana/superpowers-for-my-own-workflow --skill subagent-driven-development-galangryandana
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: subagent-driven-development
Source: https://github.com/galangryandana/superpowers-for-my-own-workflow/tree/main/skills/subagent-driven-development
Command: npx skills add https://github.com/galangryandana/superpowers-for-my-own-workflow --skill subagent-driven-development-galangryandana

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill enables teams to execute implementation plans by dispatching a new subagent for each task, paired with two-stage reviews to ensure spec compliance and code quality while keeping work fast and isolated.

Core Features & Use Cases

  • Per-task subagent execution: Each task is handled by a fresh subagent to avoid context pollution and improve focus.
  • Two-stage review: After implementation, a spec-compliance review verifies alignment with requirements, followed by a code-quality review.
  • Rapid iteration in a single session: Tasks are processed in sequence within the same session to maintain momentum.
  • Questioning as needed: Subagents can ask clarifying questions before starting work to reduce rework.
  • TodoWrite traceability: Completion is recorded and tracked for visibility.

Quick Start

Dispatch a plan by issuing a task list, then use a fresh subagent per task with two-stage reviews to implement and verify each item.

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 implementation plans using multi-agent workflows without context pollution?

Multi-agent workflows execute implementation plans by dispatching a fresh subagent for each task. This isolated approach prevents context pollution and enables fast iteration within a single session.

What is the best way to automate code review and spec compliance for independent development tasks?

Automating code review for independent tasks is best handled by a two-stage review process. This verifies spec compliance first, then evaluates code quality after each subagent completes its implementation.

How does subagent-driven development handle task tracking and traceability?

Subagent-driven development maintains task traceability by recording task completion in TodoWrite. This provides visibility and tracks progress as each isolated subagent finishes its assigned work.

Can I use fresh subagents to break down a single development session into independent tasks?

Yes, you can break down a single session into independent tasks using fresh subagents. This approach maintains momentum by processing tasks in sequence while keeping work isolated and focused.

Do subagents need clarifying questions before starting implementation work?

Subagents can ask clarifying questions before starting implementation work. This questioning reduces rework by ensuring the subagent fully understands the task requirements before writing code.

What are the limitations of using fresh subagents for implementation plans?

Fresh subagents are limited to plans that can be broken into mostly independent tasks within a single session. If tasks require heavy interdependent context sharing, the isolated subagent approach may not be suitable.