Koge Kanban

A streamlined Kanban board featuring drag-and-drop management, table views, and project organization. This application requires a local MariaDB server for data persistence and supports optional local AI integration via Ollama.

Features

• Project Management: Create multiple projects/workspaces.
• Kanban Board: Drag-and-drop tasks between custom columns.
• Table View: A structured list view of all tasks.
• Media Support: Attach image links or upload images (saved to DB).
• AI Integration: Generate task descriptions and subtasks using local LLMs (Ollama).
• Customization: Customize column colors and priority settings.
• Database Driven: Data is stored in a MariaDB database.

Prerequisites

• Node.js: Version 18.0.0 or higher.
• MariaDB: Required for data storage.
• Docker (Optional): For containerized deployment.
• Ollama (Optional): For AI features. (Recommend using Qwen2.5:latest / Qwen3:latest for concistency)

Installation Options

Option 1: Docker (Recommended)

The easiest way to run Koge Kanban is using Docker Compose. This sets up the app and database automatically.

git clone https://github.com/dezuhan/koge-kanban.git
    cd koge-kanban

1. Clone the Repository

docker-compose up -d

1. Start with Docker Compose

1. Access the App

Open http://localhost:5173 in your browser.

Option 2: Local Manual Setup

Follow these steps to run the application manually on your local machine.

git clone https://github.com/dezuhan/koge-kanban.git
    cd koge-kanban

1. Clone the Repository

npm install

1. Install Dependencies

1. Setup Database (MariaDB)

Ensure MariaDB is installed and running. Create a .env file in the root directory (optional but recommended) or rely on defaults.

Example .env:

DB_HOST=localhost
    DB_USER=root
    DB_PASSWORD=your_password
    OLLAMA_HOST=http://127.0.0.1:11434

1. Setup AI (Optional)

• Install Ollama.
• Pull a model (e.g., ollama pull gemma2:2b).
• Ensure Ollama is running (ollama serve).

node server.js

1. Start the Backend Server

The server will automatically create the koge_kanban database and necessary tables.

1. Run the Frontend

In a new terminal window:

npm run dev

Open http://localhost:5173.

Private Hybrid Mode (Local Data Gateway)

You can use the hosted frontend while keeping 100% of your data stored locally.

mkdir koge-local-server
    cd koge-local-server
    npm init -y
    npm pkg set type="module"
    npm install express mariadb cors dotenv express

1. Initialize the Project

1. Create Server File

Copy the code from [server.js](https://github.com/dezuhan/Koge-Kanban/blob/main/server.js) in this repo into a new server.js file.

node server.js

1. Run the Server

1. Connect via Browser

• Navigate to koge-kanban.vercel.app.
• Allow local network access when prompted.
• Note: AI features will only work if you have Ollama running locally on port 11434.

Project Structure

Koge-kanban/
├── components/       # React UI Components
├── services/         # Database logic (API adapter)
├── index.html        # Entry point
├── index.tsx         # React root
├── App.tsx           # Main application logic
├── types.ts          # TypeScript interfaces
├── vite.config.js    # Vite configuration
└── package.json      # Dependencies and scripts