Can I vibe code Keepa?

Build a targeted Amazon price tracking microservice using Node.js, Express, Playwright, PostgreSQL, and Prisma. The goal is to track a user-defined set of ASINs (Amazon Standard Identification Numbers) and store historical pricing, seller count, and Buy Box data.

1. DATA MODEL:
Create a Prisma schema with:
- User: id, email, createdAt.
- TrackedProduct: id, asin, marketplace (e.g., 'US', 'DE'), title, currentPrice, buyBoxSeller, targetPriceDrop, createdAt.
- PriceHistory: id, trackedProductId, price, listPrice, buyBoxSeller, primeEligible, recordedAt.

2. CORE FUNCTIONALITY:
- POST /api/products: Accepts an ASIN and marketplace. Spawns a Playwright scraper instance (using stealth plugin and residential proxy settings) to fetch initial product details (title, current price, Buy Box holder) and saves to database.
- Cron Engine: Run a scheduled job every hour using 'node-cron' that queries all active TrackedProducts, executes scraping requests via rotated proxies, appends a record to PriceHistory, and updates the parent TrackedProduct record.
- Alerting: If the scraped price drops below targetPriceDrop, send a formatted webhook payload to a Discord or Slack endpoint.
- GET /api/products/:asin/history: Returns time-series JSON array of historical prices for frontend rendering using Chart.js.

3. FAILURE MODES & RESILIENCE:
- Implement automatic exponential backoff retry logic for 503/429 responses or CAPTCHA pages.
- If a scrape fails 3 consecutive times, mark product status as 'scraping_degraded' and notify admin.
- User-agent rotation and headless browser header spoofer to evade basic detection.

4. OUT OF SCOPE:
Do not attempt to build a global search across millions of un-monitored ASINs. Focus exclusively on user-added ASINs.

$ each button prefixes agent-specific run instructions · build your own product, never copy proprietary code, trademarks or designs