Automating the web is a daunting task, and it's not always perfect.
Websites and APIs change all the time, which means that we constantly need to be ready to fix issues.
While we do have advanced monitoring and alerting systems, workflows can crash from time to time.
Type of errors
We've classified hundred of errors, from generic HTTP codes to integrations-specific errors.
We're constantly (every week!) adding new custom errors, and we'll do so until... well, probably always 😅
![]() |
Warning | This type of error is generally a functional error. It usually means that there is an action required to do on your side such as update your LinkedIn cookies. Sometimes, there is simple nothing to do on your end, which requires you to ask for help on our Support. |
![]() |
Shutdown | This type of error is a technical error. It's better to retry and contact the Support if it persists. Either the automation has:
You (and us!) can't always fix this type of errors. |
Warning Errors
The following example is a pretty good one: If you expand the "Action Required" banner, you will quickly see what happened:
Here for example we need to change the API key for the Dropcontact integration as it is no valid:
Another example with inputs
We always check imported inputs so error linked to inputs are very rare. For API users, we don't check the inputs which means that if you don't follow a field that is required or how it should be formed, you can have errors like this one:
In this case, you could always modify the inputs to retry the job with inputs that failed.
Another example with integrations
If you're using an integration like LinkedIn, your cookies might become invalid.
Avoid using LinkedIn on other automation platforms and/or disconnect from your account. Our extension will reload them every 2 hours to avoid disconnections 🙂.
To solve this issue you can typically hit "Retry now" which will redirect you to the configuration where you'll be able to update your cookies 🎉
Shutdown Errors
This type of error is generally a bit more complicated and tech-oriented.
"What is this gibberish?!"
Yes, well, that's the type of errors you can either retry or contact us to fix.
All inputs failed
This one might be tricky, though you always have the full detail.
Here's how to read them:
When you see this, you should click on the job card to be redirected to this:
Click on the only step with have to display the side-navigation menu and click on the first job:
Click on the Inputs tab:
You will be able to inspect the input to try and understand what happened.
In some cases, you will be able to retry the job by correcting the inputs.
Retry Jobs
Always check the error message before retrying.
There's no golden rules when it comes to retrying a job, you need to analyze it to understand why it failed.
Let's take the following example:
This is the kind of job you can retry, because there's an error in the parameters you provided, so it's something you'll be able to edit!
You can either click on "Retry Now" or click on the job and then on the step:
Top open the navigation menu to display the error:
And... click "Fix & resume this job to continue the workflow"; we even provide a quick "how-to" to help you correct the error 😇
Relaunch from scratch
If you feel like it's going to be easier to relaunch from scratch, you can also hit "Show configuration":
From the configuration you'll be able to "Use config" to relaunch it from scratch: