Error Monitor → Telegram (xDLx Ops)

I built this as the ops backstop for every automation I run: a single n8n Error Trigger catches failures from any workflow on the instance, and a Code node formats the payload into a clean Telegram alert with the client name (resolved from a workflow-ID lookup), failed node, error type, message, and a deep link to the execution. It also emits a structured record with a severity tag (NodeOperationError maps to high) so failures can be logged downstream to Airtable or Supabase. The result is that I hear about a broken reminder or booking flow before the client ever does.

◆Instance-wide failure alerts with per-client routing + severity tagging