subagent-driven-development

Delegate each task to a fresh subagent and run two-stage reviews.

8|1|Updated Mar 29, 2026
One-click install
npx skills add https://github.com/dxc-danny/superpowers-cn --skill subagent-driven-development-dxc-danny
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: subagent-driven-development
Source: https://github.com/dxc-danny/superpowers-cn/tree/main/skills/subagent-driven-development
Command: npx skills add https://github.com/dxc-danny/superpowers-cn --skill subagent-driven-development-dxc-danny

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

The Skill enables orchestration of complex plans by delegating each task to a fresh subagent with isolated context, ensuring clean boundaries and rapid iteration without context bleed.

Core Features & Use Cases

  • Fresh subagents per task to prevent cross-contamination.
  • Enforced two-stage reviews: spec compliance first, then code quality.
  • Parallelizable workflow design with gated handoffs for faster delivery.
  • Use case: large feature work with many independent tasks that must stay aligned with a master plan.

Quick Start

Assign each task to a new subagent and run two-stage reviews until all tasks pass quality gates.

Frequently Asked Questions about subagent-driven-development

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I delegate tasks to subagents for parallel work without context bleed?▼

To delegate tasks without context bleed, assign each task to a fresh subagent with isolated context. This enforces clean boundaries and enables parallelizable workflow design for rapid iteration on large features.

What is the two-stage review flow for subagent code generation?▼

The two-stage review flow is a gated quality assurance process that checks spec compliance first, followed by code quality. It ensures generated code meets master plan requirements before passing delivery gates.

How do I manage task lifecycles when orchestrating complex development plans?▼

Manage task lifecycles by using standardized prompts to assign tasks to fresh subagents. This orchestration method enforces per-task isolation and sequential spec and quality checks until all tasks pass.

Can I use subagent orchestration for large feature work with many independent tasks?▼

Yes, subagent orchestration suits large feature work with many independent tasks. It delegates each task to an isolated subagent, keeping them aligned with a master plan through gated handoffs and reviews.

Why should I use isolated subagents for code review instead of a single context?▼

Using isolated subagents prevents cross-contamination of context during code review and task execution. This approach enforces per-task isolation, enabling rapid QA and reliable delivery without context bleed.

When should I not use parallelizable subagent workflows for software engineering?▼

Avoid parallelizable subagent workflows for tasks lacking clear independence or spec boundaries. The orchestration relies on per-task isolation and gated handoffs, which requires tasks to be distinctly separable to function properly.