AutoTeam 🤖

---

AutoTeam is a platform-agnostic orchestration system that connects AI agents with any service through MCP servers. Think of it as an MCP hub that enables intelligent workflows across platforms, databases, APIs, and services.

🎯 What is AutoTeam?

AutoTeam orchestrates AI agents (Claude Code, Gemini CLI, Qwen Code, and more) to work autonomously across any platform that supports MCP. The agent list is fully extensible - add any AI tool that fits your needs. Instead of building custom integrations, you configure MCP servers and let intelligent agents handle complex, multi-platform workflows.

Why AutoTeam?

• 🚀 10x Productivity: Teams report handling 5-10x more routine tasks
• 🔗 Universal Integration: Connect any MCP-enabled service without custom code
• 🤝 True Collaboration: AI agents work in parallel, like real team members
• 📈 Scalable Architecture: Add agents and services as your needs grow
• 🛡️ Enterprise Ready: Container-native with full security isolation

👥 Scale Your Team with Virtual Workers

Transform your development workflow by adding AI agents as virtual team members. Each agent specializes in different roles and works in parallel, dramatically scaling your team’s capacity:

graph TB
    subgraph "Virtual Development Team"
        SD[👨‍💻 Senior DeveloperClaude Code AgentCode reviews, Implementation]
        ARCH[🏗️ ArchitectClaude Code AgentDesign, Technical decisions]
        QA[🧪 QA AssistantQwen Code AgentTesting, Quality checks]
    end
    
    subgraph "Parallel Execution"
        SD -.->|Simultaneously| FLOW1[PR ReviewsFeature Implementation]
        ARCH -.->|Simultaneously| FLOW2[Architecture ReviewTechnical Planning]
        QA -.->|Simultaneously| FLOW3[Test AutomationQuality Reports]
    end
    
    subgraph "Shared MCP Services"
        GitHub_MCP[🐙 GitHub MCPIssues, PRs, Code]
        Slack_MCP[💬 Slack MCPCommunications]
        DB_MCP[🗄️ Database MCPAnalytics, Metrics]
    end
    
    subgraph "Platform Integration"
        GitHub[GitHub Repository]
        Slack[Team Slack]
        Analytics[(Analytics DB)]
    end
    
    SD --> GitHub_MCP
    ARCH --> GitHub_MCP
    QA --> GitHub_MCP
    
    SD --> Slack_MCP
    ARCH --> Slack_MCP
    QA --> DB_MCP
    
    GitHub_MCP --> GitHub
    Slack_MCP --> Slack
    DB_MCP --> Analytics
    
    classDef agent fill:#e3f2fd,stroke:#1976d2,stroke-width:2px
    classDef mcp fill:#f3e5f5,stroke:#7b1fa2,stroke-width:2px
    classDef platform fill:#e8f5e8,stroke-width:2px
    
    class SD,ARCH,QA agent
    class GitHub_MCP,Slack_MCP,DB_MCP mcp
    class GitHub,Slack,Analytics platform

Real Impact: Teams report handling 5-10x more routine tasks with virtual workers, allowing humans to focus on strategy and complex problem-solving.

📣 Marketing Team Automation

AutoTeam also scales non-technical teams. Here’s how a marketing team leverages AI agents for content creation, campaign management, and analytics:

graph TB
    subgraph "Virtual Marketing Team"
        CM[📝 Content ManagerClaude Code AgentBlog posts, Social content]
        SM[📱 Social Media ManagerGemini CLI AgentScheduling, Engagement]
        DA[📊 Data AnalystQwen Code AgentAnalytics, Reports]
    end
    
    subgraph "Parallel Marketing Operations"
        CM -.->|Simultaneously| MFLOW1[Content CreationSEO Optimization]
        SM -.->|Simultaneously| MFLOW2[Social PostingCommunity Management]
        DA -.->|Simultaneously| MFLOW3[Campaign AnalysisPerformance Reports]
    end
    
    subgraph "Marketing MCP Services"
        CMS_MCP[📄 CMS MCPWordPress, Ghost]
        Social_MCP[📲 Social MCPTwitter, LinkedIn]
        Analytics_MCP[📈 Analytics MCPGoogle Analytics, HubSpot]
    end
    
    subgraph "Marketing Platforms"
        CMS[Content Management]
        SocialPlatforms[Social Networks]
        MarketingTools[Analytics & CRM]
    end
    
    CM --> CMS_MCP
    SM --> Social_MCP
    DA --> Analytics_MCP
    
    CM --> Social_MCP
    SM --> Analytics_MCP
    DA --> CMS_MCP
    
    CMS_MCP --> CMS
    Social_MCP --> SocialPlatforms
    Analytics_MCP --> MarketingTools
    
    classDef agent fill:#fff3e0,stroke:#f57c00,stroke-width:2px
    classDef mcp fill:#e8f5e8,stroke:#4caf50,stroke-width:2px
    classDef platform fill:#fce4ec,stroke:#e91e63,stroke-width:2px
    
    class CM,SM,DA agent
    class CMS_MCP,Social_MCP,Analytics_MCP mcp
    class CMS,SocialPlatforms,MarketingTools platform

Marketing Results: Content production increased 400%, social engagement up 250%, with data-driven insights delivered daily instead of monthly.

🎧 Customer Support Team Automation

Scale your support operations with AI agents that handle multiple channels simultaneously, ensuring no customer request goes unnoticed:

graph TB
    subgraph "Virtual Support Team"
        SC[🎧 Support CoordinatorClaude Code AgentTicket triage, Escalation]
        CR[💬 Chat RepresentativeGemini CLI AgentLive chat, Quick responses]
        KB[📚 Knowledge SpecialistQwen Code AgentDocumentation, Solutions]
    end
    
    subgraph "Parallel Support Operations"
        SC -.->|Simultaneously| SFLOW1[Ticket RoutingPriority Assignment]
        CR -.->|Simultaneously| SFLOW2[Customer ChatIssue Resolution]
        KB -.->|Simultaneously| SFLOW3[Solution ResearchKB Updates]
    end
    
    subgraph "Support MCP Services"
        Ticket_MCP[🎫 Ticketing MCPZendesk, Freshdesk]
        Chat_MCP[💭 Chat MCPIntercom, LiveChat]
        KB_MCP[📖 Knowledge MCPConfluence, Notion]
    end
    
    subgraph "Support Platforms"
        HelpDesk[Help Desk System]
        ChatPlatform[Live Chat Platform]
        KnowledgeBase[Knowledge Base]
    end
    
    SC --> Ticket_MCP
    CR --> Chat_MCP
    KB --> KB_MCP
    
    SC --> Chat_MCP
    CR --> KB_MCP
    KB --> Ticket_MCP
    
    Ticket_MCP --> HelpDesk
    Chat_MCP --> ChatPlatform
    KB_MCP --> KnowledgeBase
    
    classDef agent fill:#e1f5fe,stroke:#0277bd,stroke-width:2px
    classDef mcp fill:#f1f8e9,stroke:#689f38,stroke-width:2px
    classDef platform fill:#fef7e0,stroke:#ffa000,stroke-width:2px
    
    class SC,CR,KB agent
    class Ticket_MCP,Chat_MCP,KB_MCP mcp
    class HelpDesk,ChatPlatform,KnowledgeBase platform

Support Results: 60% faster response times, 45% better escalation accuracy, 24/7 coverage with consistent service quality across all channels.

graph TB
    subgraph "AutoTeam Core"
        ATC[AutoTeam Orchestrator]
        FE[Flow Engine]
        WM[Worker Manager]
    end
    
    subgraph "AI Agents (Scalable)"
        Claude[Claude Code Agent]
        Gemini[Gemini CLI Agent]
        Qwen[Qwen Code Agent]
        More[...More AI Agents]
    end
    
    subgraph "MCP Servers"
        GMCP[GitHub MCP]
        SMCP[Slack MCP]
        DMCP[Database MCP]
        FMCP[Filesystem MCP]
        CMCP[Custom MCP]
    end
    
    subgraph "External Platforms"
        GitHub[GitHub API]
        Slack[Slack API]
        Database[(Database)]
        FileSystem[File System]
        Custom[Custom APIs]
    end
    
    ATC --> FE
    ATC --> WM
    FE --> Claude
    FE --> Gemini
    FE --> Qwen
    
    Claude --> GMCP
    Gemini --> SMCP
    Qwen --> DMCP
    Claude --> FMCP
    Gemini --> CMCP
    
    GMCP --> GitHub
    SMCP --> Slack
    DMCP --> Database
    FMCP --> FileSystem
    CMCP --> Custom

✨ Key Features

🏗️ Architecture Overview

AutoTeam acts as an intelligent MCP hub, enabling seamless communication between AI agents and platforms:

graph LR
    subgraph "Flow Execution"
        F1[Collect GitHubGemini CLI]
        F2[Collect SlackClaude Code]
        F3[Collect DatabaseQwen Code]
        F4[Process All TasksClaude Code]
        
        F1 --> F4
        F2 --> F4
        F3 --> F4
    end
    
    subgraph "MCP Connectivity"
        F1 -.-> GitHub_MCP
        F2 -.-> Slack_MCP
        F3 -.-> DB_MCP
        F4 -.-> GitHub_MCP
        F4 -.-> Slack_MCP
    end
    
    GitHub_MCP --> GitHub_API[GitHub]
    Slack_MCP --> Slack_API[Slack]
    DB_MCP --> Database_API[(DB)]

🚀 Quick Start

Prerequisites

• Docker 20.10+ or Podman 3.0+
• 4GB RAM minimum (8GB recommended)
• Linux, macOS, or Windows with WSL2

1. Install

# One-line installation
curl -fsSL https://raw.githubusercontent.com/diazoxide/autoteam/main/scripts/install.sh | bash

# Or with specific version
curl -fsSL https://raw.githubusercontent.com/diazoxide/autoteam/main/scripts/install.sh | bash -s -- --version v1.0.0

2. Initialize

# Create a new AutoTeam project
autoteam init

# Or initialize with a template
autoteam init --template development-team

3. Configure

# autoteam.yaml
workers:
  - name: "AI Assistant"
    enabled: true
    prompt: "Handle tasks across platforms using available MCP tools"

settings:
  mcp_servers:
    github:
      command: /opt/autoteam/bin/github-mcp-server
      args: ["stdio"]
    slack:
      command: /opt/autoteam/bin/slack-mcp-server
      args: ["stdio"]

  flow:
    - name: process_tasks
      type: claude
      prompt: "Process tasks using MCP tools"

4. Generate Configuration (Optional)

# Generate configuration files for the runtime
autoteam generate

5. Deploy

autoteam up

6. Monitor (Optional)

# Access control plane API at http://localhost:9090
# View Swagger UI at http://localhost:9090/docs/

📚 Documentation

Getting Started

• 📖 Installation Guide - Complete setup instructions
• ⚙️ Configuration - Platform and agent configuration
• 🚀 Examples - Real-world use cases and templates

Advanced Topics

• 🔄 Flow System - Workflow definition and orchestration
• 🔌 MCP Integration - Platform connectivity guide
• 🏗️ Architecture - System design deep dive
• 🛠️ Development - Contributing and extending AutoTeam

Quick Links

• API Reference
• Control Plane API
• CLI Commands
• Troubleshooting
• FAQ

💡 Use Cases

Development Teams

• 🔍 Code Review Automation - Parallel PR reviews with multiple AI perspectives
• 🐛 Issue Management - Automatic triage, labeling, and assignment
• 🚀 CI/CD Enhancement - Intelligent build failure analysis and fixes
• 📝 Documentation Generation - Keep docs in sync with code changes

Marketing Teams

• ✍️ Content Production - Blog posts, social media, email campaigns
• 📊 Analytics Automation - Daily reports and campaign insights
• 🎯 SEO Optimization - Content analysis and improvement suggestions
• 📱 Social Media Management - Multi-platform posting and engagement

Customer Support

• 🎫 Ticket Automation - Intelligent routing and prioritization
• 💬 Multi-Channel Support - Unified response across chat, email, social
• 📚 Knowledge Base Updates - Automatic solution documentation
• 📈 Support Analytics - Performance metrics and trend analysis

Data Operations

• 🔄 ETL Pipelines - Intelligent data transformation workflows
• 📊 Report Generation - Automated insights and visualizations
• 🔍 Data Quality - Validation and anomaly detection
• 🗄️ Database Management - Schema updates and optimization

💻 Example: Multi-Platform Workflow

flow:
  # Parallel data collection
  - name: scan_github
    type: gemini
    prompt: "Collect urgent GitHub notifications"
  - name: scan_slack
    type: claude  
    prompt: "Check Slack for team mentions"
    
  # Process collected data
  - name: handle_tasks
    type: claude
    depends_on: [scan_github, scan_slack]
    prompt: "Process all collected tasks with appropriate actions"

🤝 Contributing

AutoTeam is open source and welcomes contributions!

How to Contribute

1. ⭐ Star the repository to show your support
2. 🐛 Report bugs via GitHub Issues
3. 💡 Request features in Discussions
4. 🔧 Submit pull requests - see Contributing Guide
5. 📖 Improve documentation - even typo fixes are valuable!
6. 🔌 Create MCP integrations - expand the ecosystem

Development Setup

# Clone the repository
git clone https://github.com/diazoxide/autoteam.git
cd autoteam

# Install dependencies
make deps

# Run tests
make test

# Build locally
make build

🔒 Security

For security issues, please email [email protected] instead of using the issue tracker. See our Security Policy for more details.

📄 License

MIT License - see LICENSE for details.

🙏 Acknowledgments

• Anthropic for Claude and MCP
• Google for Gemini
• Alibaba Cloud for Qwen
• All our contributors

---