tickfoundry
capture livesign inget the data →
§ research · note 003

The Polymarket tooling stack, mapped

2026-08-24ecosystemtools

Polymarket's third-party ecosystem is young enough that no directory, marketplace or app store maps it. The best map right now is a community-curated GitHub list: awesome-prediction-market-analytics-tools, maintained by Tal Harel of Predictbook — around 35 actively-maintained tools across nine categories, from trading terminals to tax journaling. Disclosure up front: TickFoundry is listed there (Research & Market Discovery), which is how we found it. The entry is accurate; nobody paid anybody.

Lists like this are usually noise. This one is worth ten minutes, because the shape of it — which categories are crowded and which are nearly empty — is a compact reading of where the niche actually is:

8Wallet & trader analyticswho holds what, who profits, who to copy
6Trading terminals & executionfaster order entry on the venue's own APIs
5Alerts & whale monitoringposition-size triggers, large-fill feeds
5Research & market discoveryfinding and understanding markets (TickFoundry sits here)
3Cross-platform odds aggregatorsPolymarket vs Kalshi vs books
3Trade copying & automationmirror a tracked wallet
2Signal & fair-valuemodel-derived prices vs market prices
1Portfolio, tax & journalingaccounting for a venue that settles on-chain

The center of gravity is other people's wallets

Thirteen of the thirty-five tools — wallet analytics plus whale monitoring — exist to watch other traders. That is not an accident of taste; it follows from what Polymarket makes visible. Settlement is on-chain, so every wallet's positions and P&L are public, and the cheapest edge to productize is "find the profitable wallets and watch them." Add the three trade-copying tools and nearly half the ecosystem is downstream of one fact: on this venue, the trader is a public dataset.

The quant end is underbuilt — and the reason is an input, not a lack of demand

At the other end, the categories a quant would reach for first are the thinnest: two signal/fair-value tools, one portfolio accounting tool, and no backtesting framework listed at all. The missing input explains it. Polymarket's order book lives off-chain — the chain shows settled fills, but the prices people actually traded against, the depth behind them, and every quote that came and went exist only in the venue's websocket stream, and only while it is being spoken. A fair-value tool without price history has nothing to calibrate on; a backtester without books has nothing to fill against. The watching-wallets half of the ecosystem is crowded because its input is free; the modeling half is empty because its input, until recently, did not exist.

That input is the business we are in — we record the book itself, tick by tick, and it is why our entry sits in the discovery section of the list rather than the trading one. Since this week the free end of that input got wider: tickfoundry.com/candles charts OHLCV for every market on the venue — trade-tape and book-midpoint candles, free at 15-minute resolution with full history, no account. If you are building in either empty category, that plus the free sample bundle is enough raw material to prototype against before spending anything.

If you build in this niche

The list takes pull requests and reads as genuinely curated — dead tools get moved out rather than left to rot. It is the closest thing this ecosystem has to infrastructure right now, and the fastest way to see, in one scroll, what has already been built before you build it again.

Every number a tool in those empty categories needs starts as an order-book event. We record all of them.
try the candle viewer →get the sample