subagent-driven-development

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

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

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Subagent-Driven Development enables executing large plans by dispatching independent subagents per task, with two-stage reviews after each task to maintain spec compliance and high code quality.

Core Features & Use Cases

  • Fresh subagent per task for isolated work and minimal context pollution
  • Two-stage review: spec compliance first, then code quality
  • Deterministic per-task workflow with clear handoffs
  • Self-review to catch issues before reporting
  • Plan-driven task extraction and parallel-friendly execution

Quick Start

Provide a complete plan, then dispatch subagents to execute tasks with two-stage reviews until all tasks pass.

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 multi-task development workflows with isolated execution?

Multi-task development workflow automation is achieved by dispatching independent subagents per task for isolated execution and minimal context pollution. Each subagent handles a single task from the plan within the same session, maintaining a clean working environment.

What is subagent-driven development and when do I need it for complex plans?

Subagent-driven development is a technique for executing large plans by dispatching independent subagents per task. You need it for complex, multi-task development workflows where tasks are largely independent and require spec compliance and code-quality reviews.

How does the two-stage code review process work for subagent task execution?

The two-stage code review process works by running a spec compliance review first, then a code-quality review after each task. The dispatched subagent iteratively fixes issues until both review stages pass acceptance criteria.

Can I execute parallel-friendly tasks within the same development session?

Yes, you can execute parallel-friendly tasks within the same session. The subagent-driven workflow extracts tasks from a complete plan and dispatches independent subagents to execute them, ensuring deterministic per-task workflow with clear handoffs.

What's the best way to ensure spec compliance during automated task planning?

The best way to ensure spec compliance is using a deterministic per-task workflow with a dedicated review stage. Subagents perform self-reviews to catch issues before reporting, followed by a formal spec compliance review with iterative fixes.