subagent-driven-development

Dispatch a fresh subagent per task with two-stage reviews.

2|Updated Apr 25, 2026
One-click install
npx skills add https://github.com/AlexiosBluffMara/mercury --skill subagent-driven-development-alexiosbluffmara
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: subagent-driven-development
Source: https://github.com/AlexiosBluffMara/mercury/tree/main/skills/software-development/subagent-driven-development
Command: npx skills add https://github.com/AlexiosBluffMara/mercury --skill subagent-driven-development-alexiosbluffmara

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Orchestrate end-to-end plan execution by dispatching a fresh subagent per task and enforcing two-stage reviews to ensure spec compliance and high-quality code.

Core Features & Use Cases

  • Fresh subagent per task to prevent context leakage and improve focus.
  • Two-stage reviews (spec compliance then code quality) to catch issues early.
  • Structured task orchestration from an implementation plan to a complete solution.
  • Ideal for complex projects with multiple independent tasks and strict quality gates.

Quick Start

Provide an implementation plan and let Subagent-Driven Development dispatch per-task subagents with two-stage reviews to yield a finished 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 automate task delegation from an implementation plan to subagents?

Automating task delegation requires parsing an implementation plan into independent tasks and dispatching a fresh subagent per task to execute them. This approach prevents context leakage and enforces structured orchestration from plan to finished solution.

What is a two-stage code review process for subagent workflow?

A two-stage code review process enforces spec compliance checks followed by code quality reviews for each delegated task. This mechanism catches specification deviations early before evaluating the overall code quality of the subagent's output.

How do I prevent context leakage when orchestrating multiple coding tasks?

Preventing context leakage involves dispatching a fresh subagent for each independent task within your implementation plan. This isolation improves focus by ensuring each subagent operates without carrying over context from previous tasks.

Does subagent-driven development work for complex projects with strict quality gates?

Subagent-driven development is ideal for complex projects with multiple independent tasks and strict quality gates. It orchestrates end-to-end plan execution by enforcing spec compliance and code quality reviews before final integration checks.

What are the limitations of using fresh subagents per task for code generation?

Using fresh subagents per task requires an existing implementation plan with multiple independent tasks to function properly. It is not suited for single-task execution or projects where tasks have heavy interdependencies that prevent isolated delegation.