subagent-driven-development

Dispatch tasks to fresh subagents with isolated context and two-stage reviews.

Updated Apr 19, 2026
One-click install
npx skills add https://github.com/gqf2008/hermez-ai --skill subagent-driven-development-gqf2008
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: subagent-driven-development
Source: https://github.com/gqf2008/hermez-ai/tree/main/skills/software-development/subagent-driven-development
Command: npx skills add https://github.com/gqf2008/hermez-ai --skill subagent-driven-development-gqf2008

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Coordinates plan execution by dispatching tasks to fresh subagents with isolated context and built-in two-stage reviews to ensure spec compliance and code quality.

Core Features & Use Cases

  • Fresh subagent per task: isolates state and reasoning to prevent context leakage.
  • Two-stage reviews: first verify spec compliance, then assess code quality.
  • Deterministic task workflow: execute each task from a complete plan with a clear todo list.
  • End-to-end integration readiness: ensures consistency across tasks before final review.

Use cases include plan-driven feature implementation where tasks are largely independent and quality and spec conformity matter.

Quick Start

Dispatch the first implementer subagent for the initial task using the full plan context.

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 plan execution with isolated subagents for code review?

Automate plan execution by dispatching each independent task to a fresh subagent with isolated context, followed by a deterministic two-stage review chain verifying spec compliance and code quality.

What is the best way to prevent context leakage when delegating tasks to subagents?

Prevent context leakage by dispatching a fresh subagent per task, isolating its state and reasoning from previous operations to ensure clean execution before the automated review chain begins.

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

The two-stage review process assigns a spec compliance reviewer and a separate code quality reviewer to evaluate each implementer subagent's output before running a final integration check across all tasks.

When should I use fresh subagents for workflow automation instead of a single context?

Use fresh subagents for plan-driven workflows featuring largely independent tasks where strict spec conformity and high code quality matter, ensuring isolated reasoning prevents cross-task state interference.

Can I run automated spec compliance and code quality checks on independent tasks?

Yes, you can run automated checks by applying a deterministic per-task lifecycle that dispatches implementer subagents and triggers sequential spec compliance and code quality reviewers before final integration.

What are the limitations of using fresh subagents for plan-driven feature implementation?

Fresh subagents require a complete plan with clear independent tasks to function properly; highly interdependent task structures may limit the effectiveness of isolated context and the automated two-stage review chain.