Does this skill require specific settings?

Yes, this skill requires the CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS flag to be enabled to allow the creation and management of persistent teammate agents.

How does the parallel review process benefit development?

Once an implementer finishes a task, both the tester and the code-reviewer evaluate the work simultaneously. This reduces bottlenecks and ensures that functionality and code quality are verified at the same time.

What is Subagent-Driven Development?

It is a workflow where Claude acts as a coordinator, delegating implementation, testing, and review tasks to a team of specialized, persistent AI subagents that work in parallel.

What roles are included in the agent team?

The team consists of an Implementer, a Tester (adversarial e2e tests), a Code-Reviewer (spec compliance), a Demo-Presenter (manual verification), and a Demo-Reviewer (client-perspective approval).

When is it best to use this skill?

Use this skill when you have a clear implementation plan where the individual tasks are mostly independent, allowing the agent team to work efficiently without blocking each other.

Subagent-Driven Development Orchestrator

Name: Subagent-Driven Development Orchestrator
Author: pedropaulovc

bypedropaulovc

0•

Herramientas de Colaboración

Orchestrates a persistent team of specialized AI subagents to execute complex implementation plans through automated TDD and parallel review cycles.

This skill transforms Claude into a project coordinator that manages a five-member specialized agent team including an implementer, tester, code-reviewer, demo-presenter, and demo-reviewer. By leveraging the experimental agent teams feature, it automates the entire software development lifecycle—from writing failing tests and implementing code to adversarial reviews and live demonstrations. It is ideal for developers handling multi-task implementation plans where tasks are largely independent, ensuring high code quality through rigorous, multi-pass validation while maintaining deep context across iterative fix cycles.

Características Principales

01Orchestrated Test-Driven Development (TDD) workflow

020 GitHub stars

03Automated multi-agent team creation with specialized roles

04Persistent agent context retention across multiple implementation rounds

05Integrated demo generation and stakeholder-perspective review

06Parallelized code review and verification for maximum throughput

Casos de Uso

01Executing complex feature roadmaps with multiple independent modules

02Automating the full PR preparation process including testing, documentation, and demo recording

03Implementing mission-critical code requiring adversarial testing and strict compliance

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add pedropaulovc/personal-marketplace subagent-driven-development

For use in Claude.ai and ChatGPT

Download Skill