DEV Community

Cover image for 🧠 I Rebuilt Tetris with an AI Brain Using Amazon Q CLI – Neon Tetris
Kshitij Gawankar
Kshitij Gawankar

Posted on

🧠 I Rebuilt Tetris with an AI Brain Using Amazon Q CLI – Neon Tetris

#amazonqdevcli #buildgameschallenge #pygame #tetris

Ever thought Tetris needed a glow-up?

I did — and thanks to Amazon Q Developer CLI, I didn’t just build a Tetris clone... I built Neon Tetris: an AI-enhanced, retro-themed version with brainpower, beauty, and a background track that hits you with nostalgia and chill.

Neon Tetris Screenshot

This is my submission for the #BuildGamesChallenge using #AmazonQDevCLI – a global event where developers reimagine classic games through the power of conversational AI.

🧩 Challenge Breakdown

🎯 Why I Chose Tetris

Tetris is the OG puzzle game — timeless, addictive, and beautifully simple. But I wanted to bring it into 2025 with a twist: AI + glowing visuals + dynamic gameplay. Tetris gave me the perfect base to explore real-time logic, visual feedback, and adaptive difficulty all in one project.

🛠️ Prompting Techniques I Discovered

Instead of micro-managing each file or feature, I went high-level with prompts like:

"Build a modular Tetris game with AI helper logic that uses heuristics to suggest optimal moves. Include ghost piece, theme switching, and dynamic difficulty based on player performance."

This let Q CLI do the heavy lifting. I followed up with focused prompts like:

  • "Add a performance tracker that monitors score and adjusts speed."
  • "Create a theme switcher with Neon, Dark, and Retro themes."
  • "Write a helper class that calculates best piece placement using heuristics."

These small conversational tweaks saved HOURS of setup and rewrites.

🤖 How AI Handled Classic Game Dev Challenges

Q CLI impressed me with:

  • Clean module separation for game logic, AI, rendering, and input
  • Automatically generated code that worked out of the box
  • Clear comments and docstrings for most generated methods

I rarely had to fight with the code — I was just refining it and adding polish.

🔁 Automation That Saved Me Time

  • Project Structure Generation: The full game skeleton (files + folders) was built in seconds
  • Heuristic AI Logic: It created a scoring system using classic Tetris metrics like holes, bumpiness, and height
  • Control Mapping: Auto-added standard controls + pause/restart/quit states
  • Theme Manager: I just said “switchable themes” and it wrote the class, handled key inputs, and applied different styles. Insane.

🎥 Gameplay Demo

Check out the full video here! 🎮👇

📺 Watch on YouTube

🎮 Why Tetris?

Tetris is the ultimate retro puzzle game. It’s simple, fast, and satisfying. But what if it evolved with AI?

With Neon Tetris, I added:

  • 🤖 AI that suggests your next best move
  • 🎭 Switchable themes (Neon, Dark, Retro)
  • 📈 Dynamic difficulty based on player skill
  • 👻 Ghost piece to preview hard drops
  • 🎵 A peaceful Tetris piano + lullaby remix soundtrack

Built using Amazon Q CLI, this was about more than just coding — it was about prompting.

💬 Prompting the Game into Existence

Here’s an example of what I told Q CLI:

“Create a modular Tetris game with an AI Helper. The AI should recommend the best position for the current piece based on heuristics like aggregate height, complete lines, holes, and bumpiness. Add a ghost piece, dynamic difficulty, theme switching, and modular file structure.”

Q CLI broke this down into a full project with:

📁 main.py – Game loop & inputs
📁 game.py, board.py, tetromino.py – Core Tetris mechanics
📁 ai_helper.py – Heuristic-based AI recommendation engine
📁 theme_manager.py – Multiple themes with switch key
📁 performance_tracker.py – Player analytics for dynamic speed
📁 renderer.py – Handles visuals, ghost piece, UI

🧠 The AI Helper Logic

The AI evaluates every possible move using the following heuristics:

  • 📏 Aggregate Height
  • 🕳️ Holes
  • 📊 Bumpiness
  • ✅ Completed Lines

It then recommends the best move by showing a ghost piece, helping players improve their strategy.

🎛️ Controls

⬅️➡️⬆️⬇️ Arrow Keys: Move & rotate
🔳 Space: Hard drop
🅰️ A: Toggle AI helper
🔤 T: Change theme
🔍 G: Toggle ghost piece
⏸️ P: Pause
🔄 R: Restart (on game over)
❌ Q: Quit (on game over)

🧵 Final Thoughts
This challenge showed me that AI is no longer just a helper — it’s a real dev partner.
Amazon Q CLI let me build something playable, clean, and smart — just by talking to it.

Big thanks to the AWS Build Games Challenge team for making this possible.
And if you're a dev who loves building games but hates boilerplate, Q CLI is your new best friend.

🎮 Game on.

#BuildGamesChallenge
#AmazonQDevCLI #GameDev
#Python #PyGame #AI
#RetroGaming #MadeWithAI

📦 Try It Yourself

📁 GitHub Repository:

👉 https://github.com/Kshitij-0007/Neon_Tetris

🧰 Requirements:

  • Python 3.6+
  • Pygame 2.5.2+

⚡ Installation:


bash
git clone https://github.com/Kshitij-0007/Neon_Tetris
cd Neon_Tetris
pip install -r requirements.txt
python main.py
Enter fullscreen mode Exit fullscreen mode

Top comments (0)