subagent-driven-development

Dispatch isolated subagents for implementation tasks with two-stage spec and quality reviews.

Updated Apr 12, 2026
One-click install
npx skills add https://github.com/thisismynewfmail-ui/Monika-agent --skill subagent-driven-development-thisismynewfmail-ui
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: subagent-driven-development
Source: https://github.com/thisismynewfmail-ui/Monika-agent/tree/main/skills/software-development/subagent-driven-development
Command: npx skills add https://github.com/thisismynewfmail-ui/Monika-agent --skill subagent-driven-development-thisismynewfmail-ui

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Executes implementation plans reliably by dispatching isolated subagents per task and enforcing mandatory spec and quality reviews to prevent context pollution, scope creep, and silent regressions.

Core Features & Use Cases

  • Fresh context per task: Launch a new implementer subagent for each task to avoid accumulated state and confusion.
  • Two-stage review pipeline: Always run a spec-compliance review first, then a code-quality review before marking tasks complete.
  • TDD-first integration: Integrate test-driven development into each implementer context so tests drive implementation and prevent regressions.
  • Use case: Break a feature plan into small independent tasks (2–5 minutes each), have implementers produce code and tests, then run automated spec and quality reviewers before final integration.

Quick Start

Dispatch the skill with a complete implementation plan so each task is run by a fresh subagent, then enforce spec compliance followed by a code quality review before marking tasks complete.

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 code review and spec compliance for independent implementation tasks?

Automate code review and spec compliance by dispatching isolated subagents per task and enforcing a mandatory two-stage review pipeline. This process runs a spec-compliance check first, followed by a code-quality review, before marking any task complete.

What is the best way to prevent context pollution when executing multi-task feature plans?

The best way to prevent context pollution is launching a fresh implementer subagent for each independent task. Providing the full task text within each isolated delegate context avoids accumulated state and confusion across the implementation plan.

How do I structure a feature implementation plan for parallel subagent delegation?

Structure the plan by breaking the feature into small, independent tasks that each take 2–5 minutes to execute. Provide the complete task text for each delegate so the subagent can produce code, run tests, and undergo reviews independently.

When should I avoid using subagent-driven development for task delegation?

Avoid subagent-driven development when implementation tasks are highly dependent on each other or cannot be isolated. This approach requires mostly independent tasks to ensure fresh subagent contexts and parallel execution function without cross-task state conflicts.