subagent-driven-development

Delegate isolated tasks to fresh subagents with two-stage reviews.

Updated Aug 27, 2026
One-click install
npx skills add https://github.com/huangzida/agents --skill subagent-driven-development-huangzida
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: subagent-driven-development
Source: https://github.com/huangzida/agents/tree/main/skills/subagent-driven-development
Command: npx skills add https://github.com/huangzida/agents --skill subagent-driven-development-huangzida

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Coordinators and teams often get bogged down managing long, multi-step implementation plans within a single session. Subagent-Driven Development isolates tasks to fresh subagents, prevents context leakage, and enables rapid iteration with structured reviews.

Core Features & Use Cases

  • Fresh subagent per task with clear handoffs and two-stage reviews (spec compliance, then code quality).
  • Plan-driven execution that keeps the main session focused while subagents handle discrete work.
  • Automated self-reviews and reviewer prompts to catch issues early across tasks.
  • Use Case: when executing a multi-step feature plan where tasks are interdependent but can be executed in isolation with precise coordination.

Quick Start

Dispatch a new implementer subagent for the first task and run a two-stage review sequence after 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 isolate tasks for multi-step code execution to prevent context leakage?▼

To isolate tasks, delegate each discrete work item to a fresh subagent within a single session. This prevents context leakage by keeping the main session focused while subagents handle isolated execution with structured handoffs.

What does subagent orchestration do for long implementation plans?▼

Subagent orchestration coordinates multi-task execution by delegating isolated tasks to fresh subagents. It keeps the main session focused on coordination while subagents execute discrete work with strict context separation and deterministic progress tracking.

When should I use isolated subagents for task dispatch?▼

Use isolated subagents when executing a multi-step feature plan where tasks are interdependent but can be executed in isolation. This approach benefits plans requiring strict context separation, per-task reviews, and coordinated progress tracking.

How do I run a two-stage code review across multiple subagents?▼

Run a two-stage review sequence after each task: first check spec compliance, then evaluate code quality. Dispatch a new implementer subagent for the next task while applying automated self-reviews and reviewer prompts to catch issues early.

Does subagent-driven development work for interdependent tasks in a single session?▼

Yes, it coordinates multi-task execution within a single session by delegating isolated tasks to fresh subagents. It applies deterministic task orchestration across the workflow while maintaining strict context separation between interdependent tasks.

What are the limitations of using fresh subagents for every task?▼

Fresh subagents prevent context leakage but require clear handoffs and coordinated progress tracking. This approach suits multi-step plans with independent tasks, but tasks needing continuous shared context may not benefit from strict isolation.