Your Role in Automation Management
Owners create automations; managers maintain them. Your job is to ensure automations are running correctly, identify failures before they impact operations, and escalate to the Owner when an automation needs reconfiguring.
Viewing Active Workflows
Navigate to AI → Automations. The automations list shows:
- Name and description
- Status — Active, Paused, or Failed
- Last run — timestamp of the most recent execution
- Run count — total number of times it has fired
- Success rate — percentage of runs that completed without error
A status of Failed means the automation's last execution did not complete successfully. These need immediate attention.
Checking Run History
Click any automation name, then select the Run History tab. Each run shows:
- Timestamp — when the automation fired
- Trigger event — what caused it to fire (e.g. "Product X stock dropped to 8 units")
- Conditions evaluated — which conditions passed or failed
- Actions executed — each action in sequence with its result (Success / Failed / Skipped)
- Duration — how long the automation took to run
- Error message — if the run failed, the specific error is shown here
Sort run history by Status: Failed to quickly find problematic runs. The error message usually tells you exactly what went wrong — common failures include "Supplier email bounced", "Stock update conflict", or "Stripe API rate limit exceeded".
Pausing a Broken Automation
If an automation is failing repeatedly:
- Click the automation name
- Click Pause
- Confirm the pause — the automation stops firing immediately
- Investigate the cause using the run history error messages
- If you can fix it (e.g. update a supplier email address), do so and click Resume
- If the automation logic needs changing, click Escalate to Owner — managers cannot edit automation triggers or actions, only pause/resume them
Don't leave a failed automation running. Each failed run may create incomplete records (e.g. a draft PO with missing line items, or a notification that was partially sent). Pause it first, investigate, then resume or escalate.
Understanding Failure Logs
The Failure Log (AI → Automations → Failure Log) aggregates all failed runs across all automations in one place. It shows the failure rate trend over the past 30 days and groups failures by error type.
Common error types and what they mean:
| Error | Likely cause |
|-------|-------------|
| integration_timeout | External service (Stripe, WhatsApp) was slow or offline |
| duplicate_record | Automation tried to create a record that already exists |
| permission_denied | The action requires a higher policy risk level |
| condition_error | A condition referenced a field that no longer exists (e.g. deleted category) |
| rate_limit | Too many automations fired in a short window |
Adjusting Trigger Conditions
When the Owner has given you permission to adjust conditions (shown in the automation's settings), you can:
- Click the automation name
- Click Edit Conditions
- Modify the condition values (e.g. change the threshold quantity or date range)
- Click Save — the change takes effect immediately on the next trigger event
Note: you can only adjust condition values, not add or remove conditions. Structural changes require Owner access.
Reviewing AI-Suggested Automations
The AI may suggest new automations based on patterns it observes. These suggestions appear at AI → Automations → Suggestions. Each suggestion shows the proposed trigger, conditions, and action, along with the reasoning.
Review suggestions and forward the ones you think are valuable to the Owner with a note using the Share with Owner button.
Next Steps
The final lesson in the Manager path covers the Knowledge base and Playbooks — how to access operational memory and follow structured response procedures.