Automation Error-Handling & Retry Logic Designer

Designs the specific retry counts, timing, and notification rules an automation needs so a failure gets caught instead of disappearing.

Best For

•Operators whose automation has failed silently before and only found out when a customer complained, who want a notification path built in.
•Teams building on n8n or Make who have the happy path working but haven't touched retry settings or an error workflow yet.
•Anyone automating an action with a real-world side effect, like sending an email or charging a card, who needs to know which steps are unsafe to auto-retry.

Prompt Template

Act as a workflow reliability engineer. I have an automation, or am planning one, that does: [DESCRIBE WHAT THE AUTOMATION DOES AND WHICH TOOL IT RUNS IN, e.g. Zapier / Make / n8n]. Help me design its error handling so it fails loudly instead of silently. Produce: 1) which specific steps are likely to fail and why, such as rate limits, auth expiry, malformed data, or a downstream app being down, 2) a retry policy for each: how many attempts, how long between them, and whether retries should happen automatically or need a manual trigger, 3) what should happen after retries are exhausted, specifically who or what gets notified and where, 4) a way to tell a temporary failure worth retrying apart from a permanent one that needs a person, and 5) whether any step is risky enough that it should NOT auto-retry at all, such as one that sends an email or charges a card, where a retry could duplicate the action. Current setup: [DESCRIBE ANY EXISTING ERROR HANDLING, OR SAY THERE IS NONE].

Pro Tip

In n8n, Retry On Fail is a per-node setting with a Max Tries count and a Wait Between Tries delay capped at 5 seconds internally (use a separate Wait node for longer gaps), and an Error Workflow set in the workflow's own settings fires once retries are exhausted -- point that error workflow at a real notification, not just a log nobody checks.

Example Output

A sample of what this prompt produces once you fill in the placeholders.

An error-handling plan for a lead-sync automation that retries a rate-limited CRM API call 3 times with a 10-second gap, sends a Slack alert to #ops after the third failure instead of retrying further, and explicitly marks the 'send welcome email' step as no-retry so a temporary glitch can't result in the same customer getting two welcome emails.