Know the moment a workflow breaks, without alert spam, and retry flaky APIs properly.

Two importable n8n building blocks for anyone running workflows in production.

1. Global Error Handler: an Error Trigger workflow you attach to any workflow via Settings. It formats each failure with workflow name, failing node, error message, execution mode, and a direct execution link, then sends it to Slack and email. Built-in dedupe suppresses repeat errors for 15 minutes, and an hourly cap stops alert storms. Both are configurable.

Retry-with-Backoff sub-workflow: call it from any workflow to make an HTTP request with exponential backoff and jitter. It respects Retry-After headers, retries only on retryable statuses (429, 5xx, timeouts), and fails clearly with an attempt log when retries run out.

What's included:

- error-handler.workflow.json, retry-with-backoff.subworkflow.json, example-caller.workflow.json

README with install steps, configuration, test procedure, and known limitations

Who it's for: n8n builders, freelancers, and ops teams who maintain client or internal automations and are tired of finding failures days late.

Requirements: n8n 1.x (self-hosted or Cloud); a Slack bot token and/or SMTP credentials.

What it's not: a monitoring platform or uptime service. It covers workflow execution errors that n8n reports, not server downtime. Credentials are placeholders; no keys are included.