Sneaker Palace
An online sneaker store where you can search thousands of shoes, read and write reviews, and check out.
React · TypeScript · Redux Toolkit · Tailwind CSS · Node.js · Express · MongoDB · Stripe · Playwright
20,000+
Sneakers from 20+ brands
790+
Automated tests run on every push to main
3 browsers
End-to-end tests run in each one
01. Overview
Sneaker Palace is a full-stack e-commerce store for sneakers. You can browse and filter a large catalog, read and post reviews with photos, keep a cart, and check out with Stripe in test mode. I built it to practice the full shopping flow end to end, from product data to payment and order history.
02. Key features





03. How it's built
Getting 20,000+ sneakers
- Problem
- The sneaker data comes from a third-party API with a small monthly request limit.
- What I did
- Wrote a script that pulled the catalog brand by brand, 100 shoes per request, and saved each shoe in MongoDB, skipping ones already saved.
- Result
- 20,000+ sneakers, and the site never calls the outside API.
Adding a test suite
- Problem
- The app had no automated tests.
- What I did
- Added unit and integration tests (Vitest), back-end tests (Supertest with an in-memory MongoDB) and end-to-end tests (Playwright), all run in GitHub Actions.
- Result
- 790+ tests run on every push.
Caching without stale data
- Problem
- Data was fetched separately in many components, and cached data could go stale.
- What I did
- Moved data fetching to RTK Query, with cache invalidation tied to the logged-in user.
- Result
- Fewer duplicate requests, and correct data after login, logout and new orders.
04. My role
I built the first version by hand. For the later rework, I planned each change (RTK Query caching, route security, accessibility and the test suite), used Claude Code to help build it, and reviewed the changes in small batches and tested them by hand. I decided what to test and reviewed every test.