Jira Cloud Automation Archive API - Audit History Shows Personal User Instead of Automation for Jira

Question

Hi Community,

I'm looking for clarification on the expected behavior of Jira Automation when using a Send Web Request action to archive issues via the Jira Cloud Archive Issue API.

Scenario

We have configured a Jira Automation rule to automatically archive issues that meet the following criteria:

status = "Done"AND statusCategoryChangedDate <= -30d

The rule actor is configured as Automation for Jira, and the archive operation is performed through a Send Web Request action that invokes the Jira Cloud Archive Issue API.

Expected Behavior

Since the rule actor is set to Automation for Jira, we expected the archive activity and issue history to indicate that the action was performed by Automation for Jira.

Actual Behavior

The automation rule executes successfully, and the Automation Audit Log shows the actor as Automation for Jira. However, the archive action appears to be attributed to my personal user account in the issue history/audit trail rather than Automation for Jira.

Environment

* Jira Cloud
* Rule Actor: Automation for Jira
* Automation Action: Send Web Request
* API Used: Jira Cloud Archive Issue API

Questions

* Is this expected behavior when using Send Web Request to invoke the Archive Issue API?
* Is the user shown in the issue history determined by the credentials used in the API request rather than the automation rule actor?
* What is the recommended approach if we want archive actions to be attributed to Automation for Jira rather than an individual user account?

The archiving itself is working correctly. Our concern is solely related to the user attribution shown in the issue history and audit records.

Thank you in advance for any guidance.

Best regards,

Nihanth Kumar

Tomislav Tobijas · Accepted Answer

@nboreddy but which API endpoint and authorization token are you using within the Send Web Request? If you're using the personal API token then that's the reason why you're seeing what you're describing 👈

You could maybe explore using service accounts for this. Rebekka did a great series on those > see A Guide to Service Accounts in Atlassian Cloud - Part 1: What they are (for); you'll just have to check if they support archiving.

Cheers,

Tobi

Lucija Maric · Accepted Answer

Hi @nboreddy, Yes, this is expected behavior when using the Send web request action. Although the rule actor is set to Automation for Jira, the REST API request is authenticated using the credentials provided in the request headers. This means the user associated with the API token is shown as the actor in the issue history, rather than the automation rule actor. If you want to avoid using your personal account, I'd recommend creating a dedicated service account and using its API token for the web request. A few things to keep in mind when setting this up: * Create a dedicated service account in your Atlassian organization. * Grant it the required Jira admin or site admin global permission to archive issues. * Generate a scoped API token for that account with the required Jira write scope. * Update your web request to use the service account's credentials and the scoped API endpoint: https://api.atlassian.com/ex/jira//rest/api/3/issue/archive You can find your Cloud ID by opening: https://.atlassian.net/_edge/tenant_info For more details, check Atlassian's documentation: * Jira Cloud REST API – Archive issues * Atlassian – Service accounts and API tokens With this approach, the archive request should be attributed to the service account instead of your personal account. 🙂

Mertcan Karabaş (WERARE) · Accepted Answer

Yes, this is expected. The rule actor applies to native automation actions such as Edit or Transition. A Send web request is a plain REST call, and it authenticates with whatever you put in the Authorization header. Atlassian's bulk-archive KB uses a Base64-encoded email:API token, so Jira records the archive as done by the owner of that token, which is your account here. Without the header, the call would run anonymously and fail.

To get cleaner attribution:

* Create a dedicated service/bot user (e.g. "Jira Archiver") with the permissions needed to archive, generate an API token for it, and use that in the header instead of your personal token. The history then shows that account and not a person. It can't show "Automation for Jira", because that app user can't issue API tokens.
* Bonus: the rule stops depending on your personal token, which is visible to anyone who can view the rule.
* Keep an eye on AUTO-1562 (https://jira.atlassian.com/browse/AUTO-1562), the request for a native Archive action. A native action would run as the rule actor.https://jira.atlassian.com/browse/AUTO-1562

KB:

https://support.atlassian.com/jira/kb/bulk-archive-issues-in-jira-cloud/