PolyScope

Free Polymarket prediction-market screener. 13,963 markets indexed, top movers, highest volume, crash signals, and category filters — all in one page.

🔗 Live: https://luciferforge.github.io/polyscope/

What it shows

• 24h movers — biggest price swings across active markets
• Volume leaders — what's actually trading, not what's just listed
• Crash signals — markets down ≥15% over 24h (one_day_change <= -0.15 in build_screener.py), a proxy for "dropped from a recent high" mean-reversion setups. See the methodology discussion for honest caveats and how the historical 73% / 5,629-event number from cross-signal-data relates to (but isn't identical to) this proxy.
• Category filters — politics, crypto, sports, economics, AI, etc.
• Active markets only — closed=false Gamma API filter, daily refresh

How it's built

• Static site generator: build_screener.py calls the Polymarket Gamma API, ranks markets, and writes docs/index.html + per-market pages.
• Hosted on GitHub Pages. Zero backend, zero subscription, zero login.
• Source of truth for the screener data: same pipeline that powers api.protodex.io and the Polymarket Historical Dataset on Gumroad.

Why this exists

Polymarket's own UI optimises for individual market depth. If you want to scan the universe for opportunities — "what dropped ≥15% in the last 24h", "where is volume actually concentrated", "what political markets exist under 20¢" — you need a screener. There wasn't a free one, so this is it.

Run it yourself

git clone https://github.com/LuciferForge/polyscope.git
cd polyscope
pip install requests
python build_screener.py
open docs/index.html

No API key required. Hits the public Gamma endpoint.

Want the underlying data?

The screener shows the latest snapshot. If you want the full historical price series — 10.8M+ snapshots across 13,963 markets, 43+ days of orderbook depth, for backtesting and feature engineering — that's the Polymarket Historical Dataset ($9, SQLite + CSV).

License

MIT.