Does the skill require manual configuration for every task?

No, it features auto-detection that analyzes your prompt to determine if the task is a feature, fix, api, or refactor, then automatically composes the appropriate team.

How does this skill save context window space?

Agents communicate by reading and writing files in a dedicated scratchpad directory rather than passing full message histories, which prevents context saturation during complex tasks.

What specialized agents are included in the team?

The skill orchestrates an Analyst (handling architecture and design), a Challenger (managing peer review and QA), and an Implementer (focusing on development and testing).

Can I run this skill on systems with limited RAM?

Yes, it includes a 'Safe' mode for systems with 2-4 GB of RAM that runs agents in a strict sequential order to minimize resource consumption.

What is the role of the 'Challenger' agent?

The Challenger acts as a reviewer that identifies potential architectural flaws, security concerns, or edge cases before the Implementation phase begins.

Multi-Agent Team Orchestrator

Name: Multi-Agent Team Orchestrator
Author: atournayre

byatournayre

•

Productivity & Workflow

Orchestrates a specialized team of AI agents to manage complex development lifecycles from initial architecture to final QA.

This skill transforms Claude into a project lead that manages a specialized micro-team consisting of an Analyst, Challenger, and Implementer. By utilizing a structured workflow (Analysis, Challenge, Implementation, and QA) and a file-based communication system via a scratchpad, it efficiently tackles complex coding tasks without exhausting context limits. It dynamically scales its execution mode based on available system resources, ensuring stable performance across different environments while maintaining high standards through rigorous peer-review cycles between agents.

Key Features

018 GitHub stars

02Context-efficient communication via scratchpad file-sharing

03Automatic task-type detection for features, fixes, and refactoring

04Automated multi-agent team creation and task delegation

05Structured development phases: Analysis, Peer-Challenge, and QA

06Resource-aware execution modes (Normal vs. Safe) based on system RAM

Use Cases

01Refactoring legacy codebases with automated safety checks and peer reviews

02Implementing complex features requiring deep architectural planning and testing

03Developing new API integrations involving multiple endpoints and security considerations

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add atournayre/claude-marketplace prompt-team

For use in Claude.ai and ChatGPT

Download Skill