CartMirror

Skin AI that meets you at checkout

Comment

People don't wonder about their skin in the abstract — they wonder right before they buy. CartMirror puts YouCam AI Skin Analysis at that exact moment.

Upload a selfie at checkout. CartMirror calls Perfect Corp's YouCam AI Skin Analysis API v2.1, reads HD concern scores (acne, pores, oiliness, moisture, redness, wrinkles, dark circles, texture, and more), then scores every product in the cart:

  • Keep — strong fit for today's concerns
  • Review — possible clash (e.g. rich cream when oiliness is high)
  • Swap — better alternatives from the catalog with reasons

Why this isn't a wrapper

The value is the decision layer: translating Skin AI scores into cart-level keep/swap guidance with explicit reasons and product suggestions — the kind of check shoppers wish retailers offered before payment.

YouCam API used

  • AI Skin Analysis v2.1 — POST /s2s/v2.1/task/skin-analysis + status polling
  • File API for selfie upload when no public image URL is available
  • Auth: Bearer API key against https://yce-api-01.makeupar.com

How to run

git clone https://github.com/happy-star/cartmirror.git
cd cartmirror
cp .env.example .env   # optional: set YOUCAM_API_KEY for live calls
npm install && npm start
# open http://localhost:3847 — click Run demo or upload a selfie

Without an API key the UI runs in clearly labeled DEMO_MODE so judges can still evaluate the full experience and matching logic.

Testing instructions

See the README Quick start. After npm install && npm start, open http://localhost:3847 and click Run demo (DEMO_MODE, no API key required) or upload a selfie. Judges can evaluate the full keep / review / swap experience and matching logic without a live YouCam key.

Built with YouCam AI Skin Analysis API v2.1, Node.js, Express, and vanilla JS. Repo: https://github.com/happy-star/cartmirror

Built With

Share this project:

Updates

Submission history