Live, public, no auth. Updated every 60 seconds. If something is wrong, this page knows before the support inbox does.
Last updated: —
Every minute the dashboard's /api/public-status endpoint runs liveness probes against the critical infrastructure: PostgreSQL primary, Redis cache, QuestDB time-series, Polygon market data, the LLM router (Claude + fallbacks), broker connectivity (MEXC + IB Client Portal), and the watchlist-intersection crons. Each probe has its own SLA threshold; the worst state determines the overall ring color.
Incidents that breach SLA for > 10 minutes get logged to errors_log with severity=critical and appear in the 24-hour history at the bottom of this page once we wire it.
Most retail SaaS hide outages until customers complain. We'd rather you see "Postgres degraded · ETA 8 min" than wonder if your alerts are broken. Status drives our on-call rotation.