How do I enable this skill in Claude Code?

It requires Claude Code with the CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1 environment variable enabled, as it utilizes experimental multi-agent features.

Does it support automated testing and review?

Yes, it includes dedicated Tester and Reviewer roles that perform mandatory quality gates, iterating with the Developer agent until code meets quality and security standards.

What is the Claude Dev Team skill?

Claude Dev Team is a specialized capability for Claude Code that spawns multiple autonomous agents (like Architects and Testers) to collaborate on complex software development tasks.

What are the different operational modes?

The skill offers 'plan' for design, 'dev' for implementation, 'full' for end-to-end with an approval gate, and 'auto' for fully autonomous execution.

Can agents communicate with each other?

Yes, the architecture allows direct peer messaging between roles like Architect and PM or Developer and Tester to resolve design conflicts and fix bugs autonomously.

Claude Dev Team

Name: Claude Dev Team
Author: rube-de

byrube-de

•

Productivity & Workflow

Orchestrates a collaborative multi-agent workflow for autonomous architectural planning, development, testing, and peer-reviewed code implementation.

Claude Dev Team transforms Claude Code into a full-scale engineering department by spawning specialized agent roles such as Architects, Product Managers, Testers, and Reviewers. It manages the entire software development lifecycle through four distinct modes, allowing users to choose between structured planning, focused development, or fully autonomous end-to-end execution. By facilitating direct messaging between agents and enforcing mandatory quality gates like automated testing and security reviews, this skill ensures that complex tasks are handled with professional-grade rigor and architectural consistency.

Key Features

01Automated feedback loops between developers, testers, and peer reviewers

021 GitHub stars

03Four operational modes (Plan, Dev, Full, Auto) to match project complexity

04Standardized handoff system using plan.md as the single source of truth

05Multi-role agent orchestration including Architect, PM, Dev, Tester, and Reviewer

06Integrated technical research specialist powered by Context7 MCP

Use Cases

01End-to-end autonomous feature development with built-in QA and security reviews

02Architecting complex system designs through collaborative agent debates

03Rapid prototyping where speed and automation are prioritized over manual oversight

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add rube-de/interstate-skills claude-dev-team

For use in Claude.ai and ChatGPT

Download Skill