Module: 5/5
Lesson: 5/7
Exercises:
Module 5 | Lesson 4

Lesson 4: Version Control for Workflows

Recovering from a Bad Change

If you've broken something, here's the calm, methodical process:

  1. Don't panic. You have backups.

  2. Stop the production workflow. In Make or n8n, toggle off the scenario or workflow trigger. It won't run again until you fix it.

  3. Find the last successful execution. In the Scenario History or Executions panel, scroll back to find the last time it worked. Click on it. Review what it produced.

  4. Compare old and new. Open your backup of the last working version. Compare it, module by module, to what's currently in production. Find what changed.

  5. Fix or restore. Either:

  6. Fix the problem you see (fix the broken condition, update the AI prompt, etc.) and test
  7. Or: delete the broken workflow and re-import the last working version, then carefully make the change again

  8. Test thoroughly. Run it manually. Check the output. If you have monitoring, make sure the execution shows success.

  9. Turn it back on. Toggle the trigger back on. Resume.

The key is: you're not starting from scratch. You have a working version to fall back on. You're not trying to reverse-engineer from memory. You're comparing two versions and finding the delta.

🔒

This lesson is premium

Get full access to AI Workflows — all modules, all lessons, lifetime access.

Already purchased? Sign in to restore access.