Your Cart
Loading

n8n Production Kit (error alerts, retries, backups, uptime)

On Sale
$29.00
$29.00
Added to cart

An n8n template for anyone who runs workflows that other people depend on. It turns a plain n8n instance into one you can leave alone: you hear about problems once, with enough detail to fix them, and you have yesterday's copy of every workflow.

What you get:

1. A global error handler. Set it as the error workflow of your workflows. Each failure sends one alert by email, Slack or both: the workflow, the node, the type of error, a suggested fix and a link to the execution. The same error again within an hour is counted, not sent again, and a cap per hour stops alert floods when something loops.

2. Automatic retries. For workflows you list, failures that are usually brief (timeouts, rate limits, 5xx, network errors) are retried through the n8n API with growing waits. If the retry works you are not bothered. If they all fail you get one alert that says so. Auth errors and bad requests are never retried.

3. A daily health digest. Runs and success rate, workflows that keep failing, active workflows with no error workflow, scheduled workflows that stopped running, runs stuck waiting, refused credentials, how many retries fixed things, and monitors that are down.

4. Expiry checks. Your own list of dates (API keys, OAuth app reviews, plan renewals), domain expiry dates, TLS certificate dates, credentials that have not been changed for months, and the findings of n8n's built-in security audit. Reported on the days you choose (30, 14, 7, 3 and 1 days before by default).

5. Nightly backups. Every workflow exported through the n8n API into a zip, one JSON file per workflow plus a manifest. A file for each weekday and one per month. A short report lists new, changed, renamed and removed workflows, and warns when someone typed a key or password into a node instead of using a credential.

6. Uptime and heartbeats. URL checks every 5 minutes (status code, required text, response time). Heartbeats for your cron jobs and workflows: they ping a URL when they finish, silence or a reported failure is an alert. One alert when something goes down, one when it is back with how long it was down.

7. A status URL that answers 503 when anything is down, so a free outside uptime checker can also tell you when n8n itself is down.

8. A dedupe gate you can call from your own workflows to drop repeats: the same order, lead or webhook delivery twice. Useful in front of steps that must not run twice.

Details that save you trouble

- One Config node for everything. A typo stops the run with a list of what to fix, and you get an alert about it.

- Every scheduled workflow has a Manual trigger and a Run Options node, so you can try it now or change a setting for one run.

- Error messages are compared with numbers, ids and dates removed, so "order 4711 failed" and "order 4712 failed" count as the same problem.

- If the email API is down, the Slack alert still goes out and the error handler still finishes.

- No AI is used. Every number is counted by code from the n8n API.

- No API keys inside the workflow files. They stay in n8n credentials.

- Everything the kit remembers is in five n8n Data Tables it creates itself.

What I tested and what I didn't

Tested on self-hosted n8n 2.44.6:

- 17 unit checks on the shared code, all passing.

- An automated suite of 14 scenarios with 81 checks, all passing. It runs the real workflows inside n8n against the real n8n API. Only the email API, Slack, the domain lookup service, the monitored sites and the failing test APIs are replaced by a local mock. It covers bad settings, alerts by email and Slack, muting repeats, auth errors, a retry that fixes the failure, retries that run out, the email API being down, every digest section, all expiry levels, the first backup, a backup with no changes, a new workflow with a typed-in key, a deleted workflow, down, slow, timeout and recovery for URLs, missed and failed heartbeats, the status URL, and the dedupe gate called per item from another workflow.

- A live run with real email to our own inbox: an error alert, the health digest, the expiry check with real domain and certificate lookups, and the backup report all arrived.

Not tested yet:

- n8n Cloud, other n8n versions and queue mode with several workers.

- Docker. The README gives the standard Docker paths, but the kit was not run in a container.

- Gmail sending (those nodes ship switched off) and a real Slack workspace (tested against a mock with the same request format).

- Very busy instances: the digest reads at most 10,000 executions by default and says when it stopped early. We tested with about 50 workflows.

- Real time scales: the schedules, the hour of muting and the weekday files were triggered by hand or shortened in tests.

- A trigger that fails to start, in n8n (covered by a unit check only).

- Restoring a backup into another n8n instance. The files were checked to be valid workflow JSON.

The full test report is in TEST-REPORT.md inside the download.

Requirements

- n8n with Data Tables and the public API. Built and tested on self-hosted 2.44.6. n8n Cloud has not been tested.

- An n8n API key.

- Resend with a verified domain, a Gmail account (untested) or a Slack incoming webhook for alerts.

- For the TLS check: NODE_FUNCTION_ALLOW_BUILTIN=tls in your n8n environment. Everything else works without it.

- Your own accounts with these providers. Their fees, if any, are billed to you by them.

Included files

- prod-1-settings.json

- prod-2-error-handler.json

- prod-3-health-digest.json

- prod-4-expiry-check.json

- prod-5-backup.json

- prod-6-uptime.json

- prod-7-heartbeat-status.json

- prod-8-dedupe-gate.json

- README.md setup guide with every setting explained

- TESTING.md: a 10 minute check

- TEST-REPORT.md: what was tested and how

- LICENSE.txt

- Screenshots

Good to know

- Setup takes about 15 minutes.

- Backups contain workflows, not credentials. n8n never exports secrets over its API, so keep your encryption key and database backup safe as well.

- The zips are written on your n8n server. Add a node to copy them to S3, Drive or similar if you want them off the server.

- Only turn on automatic retries for workflows that are safe to run twice, or put the dedupe gate in front of the step that is not.

- The kit helps you notice problems sooner. It does not guarantee that every failure is caught.

License, refunds and support

This is a downloadable, self-serve software template, not a service. You import it into your own n8n and run it on your own accounts. VNCNT Labs does not host it and has no access to your data.

License: single account. Use it in your own n8n instances and in instances you run for your clients. Do not resell or redistribute the template files. Full terms are in LICENSE.txt.

Refunds within 14 days, see vncntlabs.com/refund.

Questions: hello@vncntlabs.com

Sold by FAMVII LLC dba VNCNT Labs. n8n, Resend, Gmail and Slack are trademarks of their owners. VNCNT Labs is not affiliated with them.

You will get a ZIP (1MB) file