GhostFrog started as a tiny “let’s scrape eBay consoles” idea.
It grew into a full arbitrage intelligence system with:

• multiple scraper adapters (consoles, Apple, bikes, vans, etc.)
• Postgres pipelines for listings, comps, snapshots
• ROI calculation and modelling
• alert thresholds (profit, ROI, ‘insane deal’ markers)
• Telegram alerts (siren, digest, firehose)
• classification logic for product types
• a niche-based architecture (lego, consoles, vehicles)

🧩 Why it works

Cheap flips follow patterns.
But 99% of people don’t see them.

GhostFrog collects:

• listing data
• price history
• auction time left
• expected profit
• expected ROI
• repair/resale assumptions

It then sends:

• ROI alerts
• last-hour sirens
• summaries
• custom webhooks

I also built a unified webhook receiver on my portfolio site to capture eBay events.

GhostFrog is now the backbone of my data-driven flipping experiments — and one of the best engineering projects I’ve ever built.

🔁 Try yourself

View on Github