What is an AI ensemble team?

An AI ensemble team is a group of specialized AI agents working together in a 'driver-reviewer' mob programming style to build software using diverse perspectives and consensus-based decisions.

What documentation does the Ensemble Team skill generate?

It creates agent profiles, coordinator instructions, a PROJECT.md for constraints, AGENTS.md for conventions, architectural decision records (ADRs), and a domain glossary.

Can I customize the experts in my AI team?

Yes. During the Team Composition phase, you can research, approve, swap, or remove proposed expert personas to ensure they perfectly match your technology stack and domain requirements.

How does this skill handle different project sizes?

It provides tiered presets—Full (~9 agents), Lean (~5-6 agents), and Solo-plus (~3 agents)—allowing you to balance expertise and token costs based on your project's specific needs.

Does this skill work with existing codebases?

Absolutely. It can be used to convert an existing project to a team model by generating the necessary agent definitions, coordinator instructions, and project documentation.

Ensemble Team Setup

Name: Ensemble Team Setup
Author: jwilger

byjwilger

•

Collaboration Tools

Deploys a complete AI ensemble programming team with specialized agent personas, project scaffolding, and consensus-driven workflows.

The Ensemble Team skill automates the creation of a high-functioning AI product team within your repository. It guides you through project discovery to generate a customized 'mob' of specialized agents—such as Product Managers, UI/UX Designers, and Software Engineers—inspired by real-world industry experts. By establishing structured communication protocols, architectural decision records (ADRs), and a driver-reviewer model, this skill ensures that your AI-driven development process is governed by diverse perspectives, rigorous quality gates, and continuous retrospective improvements.

Key Features

01Event-driven retrospective protocols to ensure team alignment after every major milestone.

02Tiered team presets (Full, Lean, Solo-plus) to optimize for project complexity and token usage.

03Standardized project scaffolding including domain glossaries, ADRs, and coordinator instructions.

04Integrated 'Team Formation Sessions' to establish architectural and coding conventions via consensus.

052 GitHub stars

06Automated generation of specialized AI agent profiles with distinct technical philosophies.

Use Cases

01Bootstrapping a new software project with a full-scale, expert-level AI development team.

02Implementing enterprise-grade AI governance for complex applications requiring diverse technical perspectives.

03Migrating an existing repository to a disciplined driver-reviewer mob programming workflow.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add jwilger/agent-skills ensemble-team

For use in Claude.ai and ChatGPT

Download Skill