skills.rest
will-flemmer
will-flemmerCOMMUNITY
Explore all skills by this author →
GitHub

subagent-driven-development

Dispatch fresh subagents for each development task with continuous code review.

Updated Aug 23, 2026
Search Tags:#code review#collaboration#workflow automation#TDD#AI development#subagents#task execution
One-click install
npx skills add https://github.com/will-flemmer/claude-config --skill subagent-driven-development-will-flemmer
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: subagent-driven-development
Source: https://github.com/will-flemmer/claude-config/tree/main/skills/collaboration/subagent-driven-development
Command: npx skills add https://github.com/will-flemmer/claude-config --skill subagent-driven-development-will-flemmer

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill optimizes the execution of development plans by dispatching fresh subagents for each task, integrating continuous code reviews to ensure high quality and rapid iteration within the same session, preventing context pollution and late-stage bug discovery.

Core Features & Use Cases

  • Fresh Subagent per Task: Dispatches a new subagent for each independent task, preventing context pollution and ensuring focused execution.
  • Continuous Code Review: Integrates a code review step after each task, catching issues early and maintaining quality throughout the development process.
  • TDD Integration: Subagents are guided to follow Test-Driven Development principles for each implementation task, ensuring robust code.
  • Quality Gates: Ensures critical issues are fixed before proceeding to the next task, maintaining a high standard of work.
  • Use Case: You have a complex feature broken down into several independent tasks. This skill allows you to automate the implementation of each task, with an AI code reviewer checking the work after every step, ensuring a high-quality, efficient workflow.

Quick Start

Load your development plan and create a TodoWrite list of tasks. For each task, dispatch a fresh subagent to implement it, including tests. Once the subagent reports back, dispatch a code-reviewer subagent to review the work. Address any feedback, then mark the task complete and move to the next.