Day 11: The Fun House Photo Booth 📸✨ #6120
-
Day 11: The Fun House Photo Booth 📸✨Welcome Back, AI EngineerThe Winter Festival wants to add something magical to the entrance: a Fun House Photo Booth where visitors can take selfies with festive face filters - snowflake crowns, reindeer antlers, frosty beards, sparkling effects! The festival director, Maya, has a vision: "I want people to open it on their phones, see themselves with fun filters in real-time, click a button, and boom - instant winter magic photo they can share!" But here's the thing: Maya needs this built by tomorrow, and it's a lot of work for one person. You need to:
That's where subagents come in. 🎯 Your Mission: Build The Photo Booth App with SubagentsInstead of building everything yourself, you'll delegate work to specialized subagents - like having a dev team where each person handles a specific part of the project. This teaches you: How to break down complex projects and use subagents to parallelize development work, and when to run them sequentially vs in parallel. NEED LLM CREDITS?Sign up at goose-credits.dev to get free credits for Claude Sonnet 4.5 via OpenRouter. Use the ACCESS CODE: 💡 The Subagent ApproachThink of subagents as your temporary dev team. You're the tech lead - you describe what needs to be built, and goose will create specialized subagents to handle different parts. Here's a suggested split (but you can organize it however you want!): Subagent 1: Core App Builder
Subagent 2: Filter Engineer
Optional Subagents (if you want to level up!):
🎨 What Filters Should It Have?Get creative! Here are some ideas:
🔧 How to Use SubagentsYou'll use natural language to tell goose to create subagents and divide the work. goose automatically handles the coordination and execution. New to subagents? Check out these resources to learn how they work: 📖 Resources:
🎥 Watch & Learn:
💡 Key Concepts to Understand: Sequential vs Parallel:
Think about: Does Subagent 2 need Subagent 1's work to be done first? Or can they work independently? 📝 RequirementsCore Functionality: 🎯 Level Up (Bonus Challenges)Beginner 🌟:
Intermediate 🌟🌟:
Advanced 🌟🌟🌟:
Ultimate Challenge 🌟🌟🌟🌟:
📤 Share Your Work📮 Required: Submit in the Advent of AI Discussion under Day 11. Accepted Formats:
|
Beta Was this translation helpful? Give feedback.
Replies: 11 comments 2 replies
-
|
Here's my fun house photo booth that I built with subagents and deployed on vercel! https://funhouse-photo-booth.vercel.app/ |
Beta Was this translation helpful? Give feedback.
-
Beta Was this translation helpful? Give feedback.
-
|
Day 11 ✅ TIL that subagents exist! Never came across them somehow, but it was really cool seeing them all go off at the same time in front of me and do different tasks! I got this screenshot part-way through: Ended up deploying it to a live GitHub pages site and you can see the repo here! Works best on a phone from my experience, if the camera is zoomed in, flip it to use the back camera and flip it back. Works best with the stars filter (it was pretty difficult getting the hat/antlers to match up!) |
Beta Was this translation helpful? Give feedback.
-
|
Filter Both with some of every challenge level. I wanted to see MediaPipe face overlay/mesh worked pretty good. Will deploy and update link.
MediaPipe.Face.mp4 |
Beta Was this translation helpful? Give feedback.
-
|
I got photos working but my face detection isn't working yet. I also fell asleep last night early. Time for Christmas vacation? 😅 I'm reusing what I did for hand detection in day 5 but for face. Current progress Progress here https://github.com/nickytonline/advent-of-ai-2025/tree/main/day-11/photo-booth Full repo for all says at https://github.com/nickytonline/advent-of-ai-2025 |
Beta Was this translation helpful? Give feedback.
-
|
POV: You delegate work to subagents and they actually deliver. Day 11's solution for Advent of AI is live. Advent.of.AI.-.Day.11.mp4 |
Beta Was this translation helpful? Give feedback.
-
|
I'm catching up, but my solution is live here! https://adventofai.vercel.app/day-11-photo-booth/index.html. It uses two subagents to do the work. |
Beta Was this translation helpful? Give feedback.
-
Beta Was this translation helpful? Give feedback.
-
|
Beta Was this translation helpful? Give feedback.
-
Beta Was this translation helpful? Give feedback.
-
|
I had to work with goose for about 20 minutes for it to finally work due to the server not working...This was my results of the challenge once it did work: |
Beta Was this translation helpful? Give feedback.










POV: You delegate work to subagents and they actually deliver.
Day 11's solution for Advent of AI is live.
Advent.of.AI.-.Day.11.mp4
https://www.youtube.com/shorts/JTy7cBwrdio