Vibe Coding: Built a game "Cloud Catcher" with Amazon Q CLI
Harnessing the fun of "vibe coding" to build a Pygame catch‑and‑dodge game—no code typed by me!
Published Jun 10, 2025
Over the weekend, I embarked on a creative sprint with a clear aim: use Amazon Q CLI to generate a playable game prototype—entirely via dialogue. What started as a simple "catch the falling icons" concept evolved rapidly. In just a few prompts, I had:
- Core gameplay (catch/miss mechanics)
- Scoring and life systems
- Sound, UI, polish
- Data persistence and replay options
All built through vibe coding—that interactive, sleep-no-IDE, prompt–run–tweak flow that turns development into pure fun.
Before we dive in, here’s how to get ready:
1. Installed Amazon Q CLI and authenticated with my Builder ID using:
2. Then, verified the environment using below command if everything is working:
3. Proceeded with setting up Python + Pygame:
4. And then, ready with the prompts firing with vibe-code:
My opening prompt:
Within seconds, Q CLI returned working code: window, player movement, falling objects—playable right away.
Then, ran the game using:
Next, I iterated feature-by-feature via natural language:
Prompt1:
Prompt2:
Prompt3:
Prompt4:
Prompt5:
Prompt6:
Once features were stable, I asked Q CLI for polish and performance:
- Prompt: "Add gentle star-field background animation".
- Prompt: "Optimize object spawning and off-screen cleanup for smoother 60 FPS".
- Prompt: "Make red AWS cost spike icons flash briefly when spawned".
The result? A crisp, engaging arcade feel, all built from my terminal prompts.
- Vibe coding ⚡: prompt → modify → test → repeat
- AI collaboration: Q CLI handled all boilerplate, UI updates, and fixes
- Debug support: I posted crash logs and Q CLI helped instantly
It felt like pair-programming with an AI co-pilot—interactive, fast, creative.
- Being clear & concise: e.g., "Deduct a life on cost spike" not "make it hard"
- Iterate small features vs overwhelming requests
- Always test after each prompt to verify behavior
- When bugs arose, share the error message
I’m writing this post right here on AWS Community—following the #AmazonQCLI campaign rules. Here’s what I included:
- Why Cloud Catcher? (mix of nostalgia and AWS branding)
- My step-by-step vibe coding experience
- Prompt examples + inline code snippets
- Screenshots of gameplay
- How Q CLI simplified design, iteration, debugging
Want to try Cloud Catcher yourself or build on it?
👉 Check out the full code here: Cloud Catcher
Feel free to ⭐️ the repo or fork it and vibe-code your own twist!
With Amazon Q CLI, I built Cloud Catcher—a fully featured, fun Pygame project—without writing a line of code. From prompts to polish, the whole vibe-coding workflow was engaging, empowered, and efficient.
So if you're seeking a fun, code-powered creative burst—fire up Q CLI, pick a game idea, and start vibe coding. Trust me, your terminal will thank you for it. Game on, and happy vibe-coding! 🚀