🦖Term-Rex: Dino Runner Game but in your terminal

JiaNong — Mon, 12 May 2025 06:56:22 +0000

This is a submission for the Amazon Q Developer "Quack The Code" Challenge: That's Entertainment!

What I Built

Yes. As you can see, this is the classic Chrome offline Dino Runner game, but in command-line version. I call it Term-Rex or <T/>-rex.

What makes Term-Rex special is that Amazon Q Developer handled approximately 90% of the development work.
From designing the game architecture to implementing complex game mechanics like collision detection, obstacle generation, and animation systems - Amazon Q was the driving force behind this project.

Demo (no gif here so:)

How to Play

Term-Rex can be installed and played through multiple methods:

Homebrew (macOS/Linux):

brew install jianongHe/tap/term-rex
term-rex

NPM:

npm install -g term-rex
term-rex

Scoop (Windows):

scoop bucket add jianongHe https://github.com/jianongHe/scoop-bucket.git
scoop install term-rex

Direct Download

Download the pre-built binary for your system from the releases page.

Or you can directly clone my repo and go run main.go

Game Controls:

Key	Action
Space / ↑	Jump
↓	Duck
P	Pause/Resume
R	Restart (after game over)
Q / Esc	Quit

Code Repository

https://github.com/jianongHe/Term-Rex

How I Used Amazon Q Developer

Did you notice the dino's eye is a Q? Guess what? That means Amazon Q, haha.

To be honest, 90% of this game's code was done by Amazon Q. I just described everything I wanted in the terminal, and it basically did the rest.

What impressed me most about working with Amazon Q Developer was:

Running in terminal: It can directly run commands in the terminal, which makes it way more powerful. That means it can access more system operations, helping me organize project structure and run other necessary commands much more easily.
Contextual Understanding: Amazon Q understood basically the entire codebase and could make targeted improvements to specific components without breaking others.
Problem-Solving: When I encountered issues with CI/CD pipelines or version synchronization, Amazon Q quickly diagnosed and resolved them.
Code Quality: The code Amazon Q produced was clean, well-documented, and followed best practices - no spaghetti code or technical debt.
Learning Opportunity: Working with Amazon Q taught me advanced Go programming techniques and game development patterns I wouldn't have discovered on my own.

My journey with Term-Rex showcases the transformative power of Amazon Q Developer in game development. It only took me about 24 hours to finish this game Here's how Amazon Q revolutionized my workflow:

1. Architecture Design & Project Setup

I started with a simple prompt: "Help me create a terminal-based dinosaur runner game in Go." Amazon Q immediately:

Designed a comprehensive game architecture with proper separation of concerns
Set up the project structure with all necessary files and modules
Implemented the game loop with proper timing and frame rate control
Created a clean, maintainable codebase following Go best practices:

2. Game Implementation

Amazon Q Developer implemented sophisticated game core logic that would have taken me weeks to code:

Flexible animation system that handles sprite transitions for the dinosaur and obstacles
The collision detection algorithm (running, jumping, ducking).
Added a cloud system with different cloud types moving at varying speeds
Implemented ground decorations for visual depth
Help me to design ASCII art for all game elements:

3. Cross-Platform Distribution

Amazon Q helped me make Term-Rex available everywhere:

Created installation scripts for Homebrew, NPM, and Scoop(Windows)
Set up GitHub Actions workflows for automated releases
Ensured the game runs consistently across different terminal types
It can even automatically help me push codes to github:

Conclusion

During this short development experience this weekend in this project, I discovered that I've genuinely come to enjoy using Amazon Q. Initially, I thought it was just another LLM client similar to ChatGPT or Gemini, but after getting hands-on experience, I realized it can do so much more.

With well-crafted prompts, it can fully function as a "senior developer" to assist your development process. It automatically creates necessary files, implements clean and readable object-oriented code, and even runs go vet to validate code correctness. If there are any issues, it continuously works to fix them:

I believe it significantly enhances development efficiency and reliability, especially when working with strongly-typed languages like golang.

Of course, it does have drawbacks, So far I think its biggest flaw is - the dreaded reached max tokens error, haha:

Thanks for reading! Feel free to try this classic mini-game.

Note: You don't need to be offline to play it!

 █████╗ ███╗   ███╗ █████╗ ███████╗ ██████╗ ███╗   ██╗     ██████╗ 
██╔══██╗████╗ ████║██╔══██╗╚══███╔╝██╔═══██╗████╗  ██║    ██╔═══██╗
███████║██╔████╔██║███████║  ███╔╝ ██║   ██║██╔██╗ ██║    ██║   ██║
██╔══██║██║╚██╔╝██║██╔══██║ ███╔╝  ██║   ██║██║╚██╗██║    ██║▄▄ ██║
██║  ██║██║ ╚═╝ ██║██║  ██║███████╗╚██████╔╝██║ ╚████║    ╚██████╔╝
╚═╝  ╚═╝╚═╝     ╚═╝╚═╝  ╚═╝╚══════╝ ╚═════╝ ╚═╝  ╚═══╝     ╚══▀▀═╝

DashBot: 1v1 Space Robot Sprint Battle

JiaNong — Sun, 27 Apr 2025 11:56:00 +0000

This is a submission for the Alibaba Cloud Challenge: Build a Web Game.

What I Built

I built DashBot, a 1v1 space robot sprint battle game where players can't walk, only dash forward by jetting out "waste code" to move and attack. just for fun haha^ ^

Basic rule: charge, dash, knock out, or get knocked out. Watch out for shrinking space and bouncing walls — they can save you or throw you off.

Demo

Play here: Online demo

Source Code: https://github.com/jianongHe/DashBot

Alibaba Cloud Services Implementation

Elastic Compute Service (ECS)

Used for hosting the WebSocket server that handles real-time matchmaking, player sync, and battle logic. ECS gave me full control over networking, which was essential for low-latency gameplay.
Object Storage Service (OSS)

All static assets like images, sounds, and frontend code are hosted here. Uploading was fast, and it served global users reliably.
Content Delivery Network (CDN)

Accelerated asset delivery from OSS, especially helpful for players outside China. Made loading times much faster.
Alibaba Cloud Container Registry (ACR)

Used to store Docker images for my backend services. It helped me manage builds more efficiently.

Game Development Highlights

Dashed by trashy code snippets, just for fun.
Real-Time Multiplayer: WebSocket-powered 1v1 matches.
Built in 1 Week: Fast dev with strong support from Alibaba Cloud.
Physics Collisions: Charge strength affects damage and knockback.
Inspired by BattleBots: Focused on chaotic clashes with simple charge-and-dash controls.
Theme Colors: Based on Evangelion Unit-01—purple and green for a sci-fi vibe.

Problems I Ran Into:

Ghost Collision Issue: Sometimes two players would overlap due to server delay, making collisions feel "soft" or unresponsive. Fixed by tweaking the sync rate and improving client-side prediction.
High Refresh Rate Drift: On Windows with high refresh monitors, physics started breaking. Solved by forcing frame sync.
Lag and Desync: Classic network game problems—adjusted tick rates, interpolation, and still had weird bugs.
Collision Bugs: After dashing into each other, sometimes unexpected stuff happened. Honestly, I just skipped a frame to patch it (￣▽￣)

AI Help:

I used ChatGPT/Gemini/Grok/v0.dev during development. It helped a lot by generating usable code quickly, though many parts still needed manual fine-tuning.

P.S.

This was my first time making a real-time multiplayer game, and it was tough but fun. I learned a lot about networking, sync issues, and how chaotic robot battles can be.

I discovered this challenge only one week before the deadline. It was tough to pull off, but I already had the idea in mind and managed to bring it to life. There’s still room for improvement, and I tried to polish it as much as I could within the limited time.

Thanks for checking out DashBot!

DEV Community: JiaNong

🦖Term-Rex: Dino Runner Game but in your terminal

What I Built

Demo (no gif here so:)

How to Play

Direct Download

Code Repository

How I Used Amazon Q Developer

1. Architecture Design & Project Setup

2. Game Implementation

3. Cross-Platform Distribution

Conclusion

DashBot: 1v1 Space Robot Sprint Battle

What I Built

Demo

Alibaba Cloud Services Implementation

Game Development Highlights

Problems I Ran Into:

AI Help:

P.S.