Checking systems…
Last updated: · Auto-refreshes every 5 minutes
Operational
Issues
Total checks
⚠ Active Incidents
Functional checks — each member receives a real API call every 15 minutes. Green = responding. Red = member is down and you've been emailed.
Outcome checks — verifies Make.com is delivering emails, tickets are being handled, and auto-resolve is running. Not just uptime.
Outcome checks — verifies the agent is adding prospects, sending outreach emails, and no creators are stuck unsent.
Verifies nightly page upgrades, new page generation, GA4 sync, and pixel events are all producing output.
Every scheduled job checked against its expected interval. Amber = overdue. Red = last run failed.
Outcome checks on key edge functions — Claude proxy health, Shopify order ingestion, Meta ads activity, creator reply detection, and scheduled dispatches.
Architecture
Status checks are performed server-side by a Supabase edge function (status-monitor) running in the same region as the RB infrastructure — not from your browser. This eliminates CORS false positives that plague client-side status pages.

The monitor runs on a pg_cron schedule every 15 minutes. Results are stored in rb_status_checks in Supabase. This page fetches the latest results from that function on load and every 5 minutes.
What Each Check Actually Does
Netlify sites — HTTP HEAD request. Green = 2xx/3xx response. Red = 5xx or timeout.
Supabase / Resend / Shopify / Anthropic — Parses official status JSON APIs. Reads the indicator field directly.
CS Pipeline — Checks if any ticket arrived in last 24h (Make.com health), if open tickets are unhandled for 2h+, and if auto-resolve ran last night.
Influencer Pipeline — Checks if the agent added prospects in last 25h, if outreach emails were actually sent, and if any creators are stuck unsent for 48h+.
SEO Pipeline — Queries page_upgrade_queue and generated_pages_log to confirm nightly jobs produced output.
Analytics — Checks ga4_daily for last sync date and analytics_events for recent pixel activity.
Cron Jobs — Every scheduled job checked against its expected run interval. Flags overdue or failed jobs immediately.
Edge Functions — Outcome checks: Claude proxy pinged directly, Shopify orders verified in database, Meta ads activity checked, creator replies and monthly/weekly dispatches verified against actual run history.
When a Break Is Detected
Critical failures (CS pipeline down, influencer agent silent, Claude API down) trigger an alert email immediately on first detection — no delay.

Warnings (degraded services, overdue crons) alert when a check moves from green to amber for the first time.

Alerts go to saul@rachelbloom.com only, naming exactly what failed and why. A recovery email is sent automatically when the service comes back online.
Status Indicators
Operational — responding normally
Degraded — responding but with issues (e.g. Shopify reports a minor incident)
Down — not responding or reporting a major incident

The hero bar at the top reflects the aggregate state. If 3 or more services are down simultaneously it turns red and reads "Service Disruption Detected."