How to tell if your automations are fragile

Fragile automations don't announce themselves. They work — until they don't. And when they break, they tend to break at the worst possible moment.

Here's how to spot fragility before it becomes a crisis.

No one knows how it works

If the answer to "how does this automation work?" is "ask [specific person]" — that's fragility. Systems that depend on individual memory are one departure away from becoming unmaintainable.

It's been "working" without anyone checking

Silence isn't stability. If no one has looked at the workflow in months, you don't know if it's working correctly — you just know it's running.

Fragile systems often produce subtly wrong outputs for a long time before anyone notices.

Every fix creates a new problem

When fixing one thing breaks another, you're dealing with tight coupling. The system has grown beyond anyone's ability to reason about it.

This is usually a sign the automation was built incrementally without design — bolt-on after bolt-on.

There's no error handling

What happens when an API call fails? When data is missing? When a user enters something unexpected?

If the answer is "the whole thing breaks" or "I don't know" — that's fragility.

The checklist

• Can someone new understand it from documentation alone?
• Is there monitoring in place?
• Do you know when it last ran successfully — and how to check?
• Has it been reviewed in the last 6 months?
• Is there a rollback plan?

If you answered "no" to more than two, the system is fragile. It's not a matter of if it will break — it's when.