Mission-Critical Watchdog
Last run pending
Production heartbeat, smoke routes, queue processing, and founder paging.
Founder Control Plane
CodePager
See what needs you now across the 30-minute watchdog, maintain lane, and urgent inbound that should not wait for a once-a-day email check.
Projects, routes, and thresholds are control-plane data. Skills and source kits are versioned separately so agents can onboard and instrument without repo archaeology.
Current stance
Healthy
Watchdog healthy. No open incidents. Public smoke routes clean.
Maintenance Lane
Separated
Maintenance jobs stay visible but never masquerade as outage monitoring.
Urgent Inbox
Contact-form triage next
Billing issues, sales questions, and account blockers should become first-class dashboard events.
Recent Incidents
Timeline
Latest resolved incidents from heartbeat state
Why This Matters
Operator Clarity
- The dashboard must show whether the pager ran recently.
- It must separate outage monitoring, maintenance, and urgent inbound.
- It must explain who got paged and why, not just show red lights.
- It must make stale monitoring itself look like an incident.