subagent-driven-development

Execute implementation plans by delegating tasks to fresh subagents with two-stage review.

Updated Aug 27, 2026
One-click install
npx skills add https://github.com/big4council-prog/b4c-agent --skill subagent-driven-development-big4council-prog
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: subagent-driven-development
Source: https://github.com/big4council-prog/b4c-agent/tree/main/skills/software-development/subagent-driven-development
Command: npx skills add https://github.com/big4council-prog/b4c-agent --skill subagent-driven-development-big4council-prog

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

It turns an implementation plan into reliable code by executing tasks with fresh subagents and enforcing two-stage validation (spec compliance first, then code quality).

Core Features & Use Cases

  • Two-stage review pipeline: Verifies each task against the original plan spec, then checks code quality and security before moving on.
  • Fresh subagent per task: Prevents context pollution by giving each task a clean, focused execution environment.
  • Plan-to-todo execution: Reads and parses a plan once, then dispatches tasks with complete task context instead of re-reading the plan repeatedly.
  • TDD-aligned implementation: Requires implementer subagents to write failing tests first, implement minimally, verify tests pass, and avoid regressions.
  • Batch integration pass: Performs a final integration review after all tasks are completed.

Examples

  • Converting an authentication or feature plan into working code by breaking it into small tasks like “create model”, “add hashing”, and “create endpoint”, with automated spec and quality checks for every task.

Quick Start

Ask the agent to execute your existing implementation plan by running the subagent-driven-development workflow with spec-compliance review followed by code-quality approval for 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 automate implementation plan execution with spec compliance and code quality checks?▼

Automated implementation plan execution works by parsing a plan into a task list, delegating each task to a fresh subagent, and enforcing two-stage gated reviews for spec compliance and code quality before proceeding.

What is the best way to prevent context pollution when delegating feature implementation tasks to subagents?▼

Preventing context pollution requires spinning up a fresh subagent for each task, giving every implementation step a clean environment and complete task context instead of reusing a single agent.

How does TDD-first execution work with parallel subagent delegation?▼

TDD-first execution requires implementer subagents to write failing tests first, implement code minimally to pass tests, verify no regressions, and then pass two-stage validation before moving on.

Do I need an existing implementation plan before starting subagent-driven development?▼

Yes, you need an existing implementation plan. The workflow reads and parses the plan once into an explicit task list, then delegates tasks with complete context rather than re-reading the plan repeatedly.

Can I use subagent-driven development for independent feature delivery tasks requiring regression checking?▼

Yes, it applies to feature delivery workflows where tasks are independent, quality and spec adherence are required, and TDD-first execution with regression checking is desired.

What happens after all implementation tasks pass spec and quality reviews?▼

After all tasks pass spec compliance and code quality verdicts, a batch integration pass performs a final integration review to verify the complete feature delivery works together.