Monitoring Rules & Auto-Rerun

A monitoring rule is a standing instruction: "when this automation (or anything in this folder) fails, recover it." Without rules, Nimbus still records failures — rules are what make recovery automatic.

Rule types #

Folder rules are convenient for teams that organize automations by client or campaign — one rule covers the whole folder, and new automations added to it are protected automatically.

Creating a rule #

From Monitoring → Rules, click Add rule:

1. Choose Automation or Folder.
2. For an automation rule, search for the automation by name — Nimbus autocompletes from automations seen in the last 30 days.
3. For a folder rule, pick the folder from the synced Automation Studio folder tree and decide whether to include subfolders.
4. Set the recovery options below.
5. Save. The rule is enabled immediately.

Recovery options #

• Honor "run once" — when on, Nimbus will not rerun an automation that SFMC has flagged as run-once, avoiding rerun attempts on automations you may be developing or testing.
• AI recovery enabled — allow a second, AI-assisted attempt if the plain rerun fails.
• AI recovery mode — Advisory, Assisted, or Autonomous. See AI-Powered Remediation.

You can change any of these later from the rule's card without recreating it.

What a rerun does #

When a rule matches a failure, Nimbus creates a rerun attempt and:

1. Waits — polls the original run until every activity has left the "Running" state.
2. Trims — builds a copy of the automation containing only the failed steps and any that never started.
3. Fires — names the copy with a NIMBUS_RERUN_ prefix and starts it.
4. Enriches — the CloudPage reports the rerun's result back to Nimbus.
5. Records — the rerun attempt is marked succeeded, failed, or skipped.

Trimming matters: a 7-step automation that failed at step 5 reruns only steps 5–7, so completed work is not repeated.

Attempt numbers #

• Attempt #1 is a plain rerun — same logic, fresh execution. Many failures are transient and clear on the first retry.
• Attempt #2 is an AI-assisted rerun, created only if attempt #1 fails *and* AI recovery is enabled on the rule.

Rerun history and cleanup #

Every attempt is listed in Monitoring → Recoveries with its status and timing. Old attempts are pruned automatically by a nightly cleanup job (default retention is 90 days).

Disabling a rule #

Toggle a rule off from its card to stop automatic recovery without losing its configuration or history. Delete it only when the automation is retired.