Project: Creative Apps - Chaos Kitchen

Track

Creative Apps (GitHub Copilot)

Project Name

Chaos Kitchen

GitHub Username

@IronNathanAlvares

Repository URL

https://github.com/IronNathanAlvares/AgentsLeague_CreativeApps_ChaosKitchen

Project Description

What does your project do?
Chaos Kitchen is basically a cooking game where you throw random ingredients together and see what crazy dishes the AI comes up with. Pick weird combos like "truffle oil + kimchi + edible gold," choose a vibe (maybe cyberpunk kitchen?), and boom - you get a full recipe with backstory, plating ideas, and actual cooking steps. It's wild.

What problem does it solve?
Ever stare at your fridge thinking "I have no idea what to make"? That's the problem. But honestly, it's more about making cooking fun again instead of feeling like homework. The mini-games thing started because I thought - why not earn points while you're bored waiting for dinner ideas? Now it's this whole progression system where you unlock premium stuff by actually playing.

What are the key features?

• AI generates recipes using Ollama Mistral running locally (so it's free and your data stays private)
• Four mini-games that are actually fun: a rhythm chopping game, flame timing challenge, memory matching, and a click-speed test
• Shop where you spend earned points on fancy ingredients and kitchen themes
• Different chef personalities that change how recipes turn out - street food vendor gives you totally different vibes than a classical chef
• Save your favorites, rate them, export to PDF when you find something worth making

Built the whole thing with Copilot helping me figure out the tricky parts. No API costs, runs entirely on your machine.

Demo Video or Screenshots

Demo Video: Project: Creative Apps - Chaos Kitchen
Screenshots of the Project: https://github.com/IronNathanAlvares/AgentsLeague_CreativeApps_ChaosKitchen/tree/main/screenshots_of_project
Copilot Evidence Screenshots: https://github.com/IronNathanAlvares/AgentsLeague_CreativeApps_ChaosKitchen/tree/main/copilot-evidence

Primary Programming Language

TypeScript/JavaScript

Key Technologies Used

• React
• TypeScript
• Vite
• TailwindCSS
• Ollama Mistral AI
• Node.js
• jsPDF
• LocalStorage API

Submission Type

Individual

Team Members

No response

Submission Requirements

• My project meets the track-specific challenge requirements
• My repository includes a comprehensive README.md with setup instructions
• My code does not contain hardcoded API keys or secrets
• I have included demo materials (video or screenshots)
• My project is my own work with proper attribution for any third-party code
• I agree to the Code of Conduct
• I have read and agree to the Disclaimer
• My submission does NOT contain any confidential, proprietary, or sensitive information
• I confirm I have the rights to submit this content and grant the necessary licenses

Quick Setup Summary

1. Clone the repository
2. Install dependencies: npm install
3. Configure environment variables (copy .env.example to .env and add your Ollama API endpoint)
4. Install Ollama and pull Mistral model: ollama pull mistral
5. Start frontend: npm run dev
6. Start backend (in new terminal): cd server && node index.js
7. Open http://localhost:5173 in your browser

Technical Highlights

• Built a complete game economy with multi-game progression system - players earn points through 4 different mini-games (rhythm, timing, memory, accuracy) and spend them in a shop to unlock premium content. Balanced point payouts 5-6x lower after initial testing showed games were too rewarding.

• Implemented local-first AI architecture using Ollama Mistral instead of cloud APIs - zero costs, complete privacy, no rate limits. Set up fallback system with mock data generation so the app works even without AI running.

• Created pixel-art cooking animations and theme-specific visual feedback system - each game has unique CSS animations and sprite-based graphics. Added dark/light mode that persists across sessions with localStorage state management.

• Designed a complex React state system managing recipe generation, game scores, shop purchases, unlock progression, and user preferences - all synchronized with localStorage for persistent gameplay across sessions without needing a database.

Challenges & Learnings

The biggest challenge was getting the mini-games to feel balanced and actually fun - not just functional. My first iteration had point payouts way too high (200-500 per game), so players could unlock everything in like 5 minutes. I had to playtest a bunch, adjust the economy, and landed on 40-100 points per game with items costing 120-600. That made progression feel rewarding without being instant.

Another tricky part was making Ollama Mistral generate consistent recipe structures. AI outputs are unpredictable - sometimes it'd return a perfect JSON-like format, other times just raw text. I learned to handle that by building a fallback parser that extracts what it can and fills in missing pieces with mock data. This taught me that when working with AI, you always need a Plan B.

The rhythm game (Ingredient Chop Challenge) was surprisingly hard to implement. Getting arrow timing to sync with animations and score multipliers took way more CSS keyframe tweaking than expected. I learned about requestAnimationFrame for smooth game loops and discovered that browser performance varies wildly - what ran smooth on my desktop lagged on older laptops, so I added a simplified rendering mode.

GitHub Copilot was honestly clutch for this. I'd describe what I wanted ("make a memory card matching game with ingredient cards") and it'd scaffold 80% of it. Then I'd iterate with feedback like "add flip animations" or "track high scores in localStorage." It saved me hours of boilerplate while letting me focus on the creative stuff - game balance, visual polish, making the chaos actually chaotic.

Contact Information

Email: nathanlalvares03@gmail.com or LinkedIn: https://www.linkedin.com/in/nathan-luis-alvares-1000061bb/

Country/Region

Dublin, Ireland

This issue is a project submission rather than a requested repository change. Review the linked repository, README.md, .env.example, server/index.js, and demo materials to understand the completed Chaos Kitchen application; the issue does not define a discrete task or acceptance criteria for a newcomer.