__  __            _        _____
                            |  \/  | ___   ___( )___   |_   _|_ ___   _____ _ __ _ __
                            | |\/| |/ _ \ / _ \// __|    | |/ _` \ \ / / _ \ '__| '_ \
                            | |  | | (_) |  __/ \__ \    | | (_| |\ V /  __/ |  | | | |
                            |_|  |_|\___/ \___| |___/    |_|\__,_| \_/ \___|_|  |_| |_|

AI Workforce Command Center
Human oversight for AI coding agents

---

Why Moe?

AI coding agents are powerful but need guardrails. Moe's Tavern provides:

• Visibility - See what AI agents are doing in a Kanban board
• Control - Approve or reject AI plans before code gets written
• Traceability - Full audit log of every action
• Flexibility - Works with Claude, GPT, and any MCP-compatible agent

"Let AI do the coding, but keep humans in the loop."

---

Features

Feature	Description
Kanban Board	Visual task management in your IDE
Plan Approval	Review AI implementation plans before execution
Multi-Agent	Run architect, worker, and reviewer agents
MCP Protocol	Standard interface for AI agent integration
Real-time Sync	Live updates via WebSocket
Activity Log	Complete audit trail with log rotation
Rails System	Define constraints AI must follow
Teams	Launch parallel agent teams within epics

---

Quick Start

Prerequisites

• Node.js 18+ - Download
• JetBrains IDE (optional) - For visual Kanban board

Installation

Windows:

git clone https://github.com/yaront1111/Moe-s-Tavern.git
cd Moe-s-Tavern
.\scripts\install-all.ps1

Mac (Homebrew):

brew tap yaront1111/moe
brew install moe

Mac / Linux (from source):

git clone https://github.com/yaront1111/Moe-s-Tavern.git
cd Moe-s-Tavern
chmod +x scripts/*.sh
./scripts/install-mac.sh

Initialize a Project

# Navigate to your project and initialize
cd /path/to/your/project
moe-daemon init

# Or specify the path explicitly
moe-daemon init --project /path/to/project --name "My Project"

This creates the .moe/ folder structure with project settings and starts the daemon. Stop it with Ctrl+C or moe-daemon stop --project /path/to/project if you only want initialization.

Run Your First Agent

# Windows
.\scripts\moe-agent.ps1 -Role architect -Project "C:\your\project"

# Mac / Linux
./scripts/moe-agent.sh --role architect --project /your/project

The agent will:

1. Connect to the daemon
2. Claim a task from the board
3. Submit a plan for your approval
4. Execute the plan step-by-step

Run a Full Team

Launch architect + worker + QA agents in parallel:

# Windows
.\scripts\moe-team.ps1 -Project "C:\your\project"

# Mac / Linux
./scripts/moe-team.sh --project /your/project

---

JetBrains Plugin Installation

The plugin provides a visual Kanban board inside your IDE (IntelliJ IDEA, PyCharm, WebStorm, etc.)

Prerequisites

• JDK 17+ - Download
• JetBrains IDE - Any 2024.1+ version

Build the Plugin

The plugin bundles the daemon, proxy, agent scripts, and role documentation. Build the daemon and proxy first:

Mac / Linux:

cd packages/moe-daemon && npm install && npm run build
cd ../moe-proxy && npm install && npm run build
cd ../../moe-jetbrains
./gradlew buildPlugin

Windows:

cd packages\moe-daemon; npm install; npm run build
cd ..\moe-proxy; npm install; npm run build
cd ..\..\moe-jetbrains
.\gradlew.bat buildPlugin

The plugin zip will be created at moe-jetbrains/build/distributions/moe-jetbrains-*.zip

Install in Your IDE

Option 1: Install from Disk (Recommended)

1. Open your JetBrains IDE
2. Go to Settings/Preferences → Plugins
3. Click the ⚙️ gear icon → Install Plugin from Disk...
4. Select the zip file from moe-jetbrains/build/distributions/
5. Restart the IDE

Option 2: Development Mode

Run the plugin in a sandbox IDE for testing:

cd moe-jetbrains
./gradlew runIde    # Mac/Linux
.\gradlew.bat runIde  # Windows

Plugin Location (Manual Install)

If you prefer to extract manually:

OS	Plugin Directory
Windows	%APPDATA%\JetBrains\<IDE><version>\plugins\
Mac	~/Library/Application Support/JetBrains/<IDE><version>/plugins/
Linux	~/.config/JetBrains/<IDE><version>/plugins/

Extract the zip contents to a moe-jetbrains folder in the plugins directory, then restart your IDE.

Using the Plugin

1. Open a project in your JetBrains IDE
2. The Moe tool window appears in the right sidebar
3. The plugin auto-starts the daemon (bundled with the plugin) and connects
4. Create epics and tasks using the toolbar buttons
5. AI agents will claim tasks and submit plans for your approval

---

Architecture

graph LR
    subgraph IDE
        Plugin[JetBrains Plugin]
    end

    subgraph Backend
        Daemon[Moe Daemon]
        Files[.moe/ files]
    end

    subgraph Agents
        Claude[Claude Code]
        GPT[GPT Agent]
        Other[Other MCP Agents]
    end

    Plugin <-->|WebSocket| Daemon
    Daemon <-->|Read/Write| Files
    Claude <-->|MCP| Daemon
    GPT <-->|MCP| Daemon
    Other <-->|MCP| Daemon

Loading

Key Principle: The .moe/ folder is the source of truth. The daemon is the sole writer. All clients (plugin, agents) communicate through the daemon.

---

Project Structure

moe/
├── packages/
│   ├── moe-daemon/      # Node.js daemon (TypeScript)
│   └── moe-proxy/       # MCP stdio proxy for agents
├── moe-jetbrains/       # JetBrains IDE plugin (Kotlin)
├── docs/                # Documentation
│   ├── ARCHITECTURE.md  # System design
│   ├── MCP_SERVER.md    # MCP tool reference
│   ├── SCHEMA.md        # Data schema
│   ├── DEVELOPMENT.md   # Dev guide
│   ├── TROUBLESHOOTING.md # Common issues & solutions
│   └── roles/           # Agent role guides (bundled with plugin)
│       ├── architect.md
│       ├── worker.md
│       └── qa.md
└── scripts/             # Agent launcher & install scripts
    ├── moe-agent.ps1    # Windows agent launcher
    ├── moe-agent.sh     # Mac/Linux agent launcher
    └── install-all.ps1  # Windows full install

---

Agent Roles

Role	Purpose	Claims Tasks In
Architect	Creates implementation plans	PLANNING status
Worker	Executes approved plans	WORKING status
Reviewer	QA and code review	REVIEW status

---

Approval Modes

Configure in .moe/project.json:

Mode	Behavior
CONTROL	Manual approval required (default)
SPEED	Auto-approve after delay (configurable)
TURBO	Instant auto-approve

---

Documentation

• Architecture Overview
• MCP Server API
• Data Schema
• Development Guide
• Troubleshooting

---

Contributing

Contributions are welcome! Please:

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing)
3. Commit your changes (git commit -m 'Add amazing feature')
4. Push to the branch (git push origin feature/amazing)
5. Open a Pull Request

See CONTRIBUTING.md for detailed guidelines.

---

License

MIT License - see LICENSE for details.

---

"Welcome to Moe's Tavern - where AI agents come to get their work done!"