Until and unless we don't get a token/activity that says that the user is not using this app anymore, it's marked as active. It can occur through SSO when the user departs from the organisation, and their data is deleted. The login/auth tokens expire for the specific application.

Additionally, we can mark the user app status as inactive when offboarding them through workflows to an application that is not directly integrated.

What is the threshold for this? For example, do we mark users as not in use if we don't get token activity for 30/60/90 days?

For Okta and Azure, every user added to an application gives this status automatically for apps. So, it's basically the direct source for user-app status.

For Google WorkSpace - It was seven days. If we don't get a token, we need to verify this.

For Agents, it was seven days. We used to mark user app status inactive if data is not received for more than 7 days.