Where Polytrack data comes from—and where it can fail.

The Data API is the main public source for recent trades and wallet-position context. Polytrack uses available trade fields to build monitored records and uses position or activity responses in wallet research views. Source availability, pagination, limits, later corrections, and the timing of a request can change what is returned.

A successful response is still a snapshot. A missing row is reported as missing or unavailable context, not as proof that a wallet never traded or held a position.

Gamma supplies market discovery and descriptive metadata such as titles, slugs, outcomes, categories, resolution context, and public profile fields when available. Polytrack uses those fields to make an execution understandable. Labels can change upstream, and a public profile name remains a pseudonym rather than verified legal identity.

CLOB HTTP and websocket surfaces provide market, token, orderbook, best-price, and last-trade context. The crypto cockpit also consumes Polymarket live-data messages for short-window price observations. Connections can close, messages can arrive late or out of order, and a displayed best bid or ask is an observation rather than an assurance of an executable quote.

Polytrack bounds and validates incoming values before using them. Stale or incomplete live feeds are surfaced as degraded state instead of being silently presented as current.

After ingestion, Polytrack validates and stores records used by its feeds, severity views, analytics, leaderboard, history, heatmap, watchlists, and Telegram fan-out. Fields such as severity, alert reasons, short-window volume, baseline ratios, directional concentration, coordinated-wallet context, and new-wallet flags are Polytrack-derived classifications.

Any aggregate labeled as tracked, monitored, or flagged is calculated from this stored subset. It should not be compared with an exchange-wide statistic unless both scopes and time boundaries are made equivalent.

Signed-in users can provide watchlist addresses and private labels, star records, and connect a Telegram bot token and chat ID. Those settings control personal dashboard and alert behavior; they are not public research data and are excluded from indexable pages.

Telegram receives eligible messages through its Bot API. News links are collected from publisher RSS feeds and retain publisher attribution. These external systems have their own availability, retention, and correction behavior.

A source timestamp describes the upstream event when available. An observation or processing timestamp describes when Polytrack saw or stored it. A delivery timestamp describes a later notification attempt. These are different events and should not be collapsed into one promise of real-time delivery.

When sources disagree, Polytrack keeps the scope visible and avoids filling a missing fact with a mock value in production. Public dynamic data will require a displayed source, an as-of time, caching rules, and a failure state before it can be indexed.