Objective
Protect users from stale data and uncontrolled execution by combining feed governance with policy guardrails.
Market Data and Risk Controls
ProductionReal-time data feeds and pre-trade safety checks are treated as first-class production systems.
Market Data Channels
Ticker, orderbook, trades, and candlestick feeds are normalized in a backend-for-frontend layer for terminal consumption.
Quote tab separation allows stablecoin preview markets (iUSD) while production trading can continue on USDT reference liquidity.
Risk Gates
Order-side checks include notional limits, size boundaries, and connector availability before dispatch.
Policies should enforce explicit maximum slippage, per-strategy budget caps, and kill-switch pathways.
Operational Controls
Polling cadence and cache behavior are tuned to avoid stale front-end state under high-volume market endpoints.
Observability tracks stale-feed rate, route failure ratio, and rejected-order causes for continuous hardening.
Production Checkpoints
- ✓Kline/ticker/orderbook data refresh remains within target polling windows
- ✓Stale data alerts trigger before user-visible degradation
- ✓Pre-trade policy rejects are explicit and logged
- ✓Kill switch can halt agent and manual route dispatch
Continue exploring