🤖 Claude Team

Multi-Agent MCP Server for AI-Powered Development Teams

Orchestrate GPT, Claude, Gemini and more to collaborate on complex tasks

✨ Features

Feature	Description
🤖 Multi-Model Collaboration	Configure multiple AI models to work together, each leveraging their strengths
🧠 Smart Task Distribution	Tech Lead analyzes tasks and automatically assigns them to the best-suited experts
🔗 Workflow Templates	5 pre-built workflows: code generation, bug fixing, refactoring, review, documentation
🎯 Custom Experts	Define your own experts (Rust, K8s, Security, etc.) via environment variables
📊 Observability	Dashboard, cost estimation, and task planning preview
🌐 Proxy API Support	Custom Base URLs, compatible with various proxy services
📝 Collaboration History	Complete record of all collaborations with search support

🚀 Quick Start

Installation

hljs language-bash

# Global install
npm install -g claude-team

# Or use directly with npx (no install needed)
npx claude-team

Basic Configuration

Add to your IDE's MCP configuration file:

📍 Configuration File Locations

IDE	Path
Claude Code	`~/.claude/config.json`
Windsurf	`~/.codeium/windsurf/mcp_config.json`
Cursor	`~/.cursor/mcp.json`

hljs language-json

{
  "mcpServers": {
    "claude-team": {
      "command": "npx",
      "args": ["-y", "claude-team"],
      "env": {
        "CLAUDE_TEAM_MAIN_KEY": "sk-your-api-key",
        "CLAUDE_TEAM_MAIN_URL": "https://api.openai.com/v1",
        "CLAUDE_TEAM_MAIN_MODEL": "gpt-4o",
        "CLAUDE_TEAM_MAIN_PROVIDER": "openai"
      }
    }
  }
}

Start Using

hljs language-csharp

> Help me build a user login feature with the team

> Have the team optimize this code for performance

🎬 How It Works

hljs language-yaml

User: "Optimize this SQL query for performance"

Tech Lead Analysis →
├── Creates: SQL Optimization Expert (powerful)
├── Creates: Index Analysis Expert (balanced)  
└── Workflow: sequential

hljs language-yaml

User: "Build a settings page with dark mode"

Tech Lead Analysis →
├── Creates: UI Component Expert (balanced)
├── Creates: Theme System Expert (fast)
├── Creates: State Management Expert (balanced)
└── Workflow: parallel → review

🛠️ Available Tools

Core Tools

Tool	Description
`team_work`	🚀 Team collaboration (auto-creates experts)
`ask_expert`	💬 Consult an expert (supports custom experts)
`code_review`	🔍 Code review
`fix_bug`	🐛 Bug fixing

Workflow Tools

Tool	Description
`list_workflows`	📋 List all workflow templates
`run_workflow`	▶️ Execute a specific workflow
`suggest_workflow`	💡 Auto-recommend workflow based on task

Pre-built Workflows:

Workflow	Purpose	Steps
`code-generation`	Generate code from requirements	Design → Implement → Test → Review
`bug-fix`	Diagnose and fix bugs	Diagnose → Fix → Verify
`refactoring`	Code refactoring	Analyze → Plan → Execute → Review
`code-review`	Multi-dimensional review	Security / Quality / Performance (parallel)
`documentation`	Generate documentation	Analyze → Document

Observability Tools

Tool	Description
`team_dashboard`	🎛️ View team status, experts, models, stats
`cost_estimate`	💰 Estimate task cost (tokens, price, time)
`explain_plan`	🧠 Preview task assignment plan
`usage_stats`	📈 View model usage statistics

Integration Tools

Tool	Description
`read_project_files`	📄 Read project files for context
`analyze_project_structure`	🏗️ Analyze project structure and tech stack
`generate_commit_message`	📝 Generate commit message from diff

History Tools

Tool	Description
`history_list`	📋 View collaboration history
`history_get`	📄 Get history details
`history_search`	🔎 Search history records
`history_context`	📚 Get recent context

⚙️ Configuration

Environment Variables

Variable	Required	Description
`CLAUDE_TEAM_MAIN_KEY`	✅	Main model API Key
`CLAUDE_TEAM_MAIN_URL`	❌	Main model API URL
`CLAUDE_TEAM_MAIN_MODEL`	❌	Main model ID (default: gpt-4o)
`CLAUDE_TEAM_MAIN_PROVIDER`	❌	Provider: `openai` / `anthropic` / `gemini`
`CLAUDE_TEAM_MODEL{N}_*`	❌	Worker model N config (inherits from MAIN)
`CLAUDE_TEAM_CUSTOM_EXPERTS`	❌	Custom experts (JSON format)

N = 1, 2, 3... supports up to 10 worker models

Custom Experts

Define your own experts beyond the built-in frontend, backend, qa:

hljs language-json

{
  "env": {
    "CLAUDE_TEAM_CUSTOM_EXPERTS": "{\"rust\":{\"name\":\"Rust Expert\",\"prompt\":\"You are a Rust expert...\",\"tier\":\"powerful\"},\"k8s\":{\"name\":\"K8s Expert\",\"prompt\":\"You are a Kubernetes expert...\",\"tier\":\"balanced\"}}"
  }
}

Field	Required	Description
`name`	✅	Expert display name
`prompt`	✅	Expert role description (System Prompt)
`tier`	❌	Model tier: `fast` / `balanced` / `powerful`
`skills`	❌	Skill tags array

Model Tiers

Tier	Use Case	Example Scenarios
`fast`	Simple, quick tasks	Formatting, simple queries, docs
`balanced`	Regular dev tasks	Components, APIs, unit tests
`powerful`	Complex reasoning	Architecture, optimization, security

📦 Changelog

v0.4.0

🎯 Custom Experts - Define experts via environment variables
🔗 Workflow Templates - 5 pre-built workflows
📊 Observability - Dashboard, cost estimation, plan preview
🔌 Integration - Project file reading, structure analysis, commit messages
💡 Smart Recommendations - Auto-suggest workflows
🧪 Test Coverage - 155 test cases

v0.3.0

🔄 Task interrupt/resume support
💬 Multi-turn expert conversations
📊 Token counting and cost estimation
📋 Expert templates (6 built-in + custom)
🔔 Webhook notifications
⚡ Exponential backoff retry
🔧 Hot config reload

Earlier versions

v0.2.x

🌊 Streaming output support
📊 Usage statistics
🎯 Model strategies
💾 Result caching
🔄 Auto model switching

v0.1.x

🎉 Initial release
🤖 Multi-model collaboration
🌐 Proxy API support

🤝 Contributing

Contributions are welcome! Please read our:

Contributing Guide
Code of Conduct
Security Policy

📄 License

MIT

⭐ Star History

⬆ Back to Top

Made with ❤️ by the community

claude-team-mcp