Slack
Incoming webhook per channel. Schema digest, schedule run failure, agent offline. Click through to the workspace.
Slack setupDocumentation updates are not noisy by default. Alerts are an opt-in layer for the events you actually want to know about: a likely breaking schema change, a failed scheduled export, an agent that went offline.
Incoming webhook per channel. Schema digest, schedule run failure, agent offline. Click through to the workspace.
Slack setupTransactional email to a list of recipients. Same digest content, formatted for inbox preview and accessibility.
Email formatEvery alert attempt is tracked with channel, status, retry count, last error, and next attempt time.
Schedule deliveryRules
A rule pairs a trigger with a destination. Triggers can be schema severity, scheduled run status, or agent health. Destinations are channels. Multiple rules can fire on the same event.
Info, Attention, or Likely breaking. Pick the threshold that wakes the on-call.
Include or exclude tables, columns, indexes, views, routines, foreign keys.
A scheduled export failed three times in a row. Surface it.
The agent missed the scheduled scan or has been disconnected for too long.
Reliability
Every dispatch is a row. If the channel is down, the cloud retries with exponential backoff. The delivery log shows status, attempts, last error, and the next attempt time.
Up to six attempts with growing delays.
Timestamp, channel, status, attempt counter, last error.
Retries keep the same event details so Slack and email recipients see a predictable alert every time.
FAQ
No. Documentation updates do not generate alerts. Subscribe explicitly when you want to be paged on schema severity, schedule failures, or agent health.
Yes. Create two rules with the same trigger and different destinations, or use separate rules for different severities.
Yes. Create two rules with the same trigger and different destinations.
Slack or email. Severity threshold and filters per rule. Retried and audited.