🎮 DeepSeek Game Room

An innovative collection of interactive games where you can watch an AI (powered by DeepSeek) solve puzzles while explaining its thought process in real-time. What makes this unique is the seamless fusion of AI reasoning with game mechanics - all in-game interactions happen through the AI's detailed thought process.

🔗 Try it out here: Deployed on Railway: https://deepseekplayground-production.up.railway.app

🎯 Current Games

🎲 Mastermind

Watch as the AI strategically cracks randomly generated color codes while walking you through its deductive reasoning process step-by-step.

Key Features:

• 🎨 Random 4-color codes using 6 vibrant colors (Red, Green, Blue, Yellow, White, Orange)
• 🤔 Detailed explanation of AI's thinking process for each guess
• ✨ Visual feedback with colored pegs
• ⚡ Real-time updates of AI's attempts and reasoning

🗺️ Maze Solver

Follow along as the AI navigates through procedurally generated mazes, sharing its pathfinding strategy in real-time.

Key Features:

• 🔄 Random 6x6 maze generation with guaranteed solutions
• 👀 Real-time visualization of AI's movement
• 🧭 Transparent exploration strategy and decision making
• ✅ Clear visual feedback for successful/failed moves

🚀 Quick Start

1. Clone the repository

git clone https://github.com/Noir97/DeepSeekPlayground.git

2. Create and activate a virtual environment (recommended)

python -m venv venv
source venv/bin/activate # On Windows: venv\Scripts\activate

3. Install dependencies

pip install -r requirements.txt

4. Set your DeepSeek API key

export DEEPSEEK_API_KEY="your-api-key-here"

5. Run the application

python app.py

Structure

├── app.py # Flask application
├── requirements.txt # Python dependencies
├── runtime.txt # Python version
├── railway.toml # Railway configuration
├── Procfile # Deployment configuration
├── games/
│ ├── mastermind_game.py
│ └── maze_game.py
├── models/
│ └── deepseek_model.py
├── static/
│ ├── js/
│ │ ├── mastermind.js
│ │ └── maze.js
│ └── style.css
└── templates/
├── base.html
├── mastermind.html
└── maze.html

Dependencies

• Flask
• OpenAI Python Client (for DeepSeek API)
• Python 3.8+

Future Games (Planned)

• Battleship: AI uses search strategy with hit/miss feedback
• Hot/Cold Treasure Hunt: Navigation with temperature-based feedback
• Lock Picking Puzzle: Pattern deduction from partial information

Contributing

Feel free to open issues or submit pull requests for improvements or bug fixes.

License

MIT License