Skip to main content
This guide includes features awaiting release. If an option is not available in your workspace, contact StackShift Support.

​
What you can do

Review automated decisions, run failures and the limits that keep automation under control. Where to find it: Settings → Automation → Classification, safeguards and reports

​
Before you start

  • Access to automation configuration or reporting.
  • At least one configured workflow or classification policy.

​
Steps

1
Define the topics your team uses to classify conversations. Review the supported languages and the active policy.
2
For AI classification, review workspace opt-in and its token budget before activation.
3
Review automation safeguards and choose limits suitable for your team.
4
Open reports or run history, select a failed or unexpected run and read its step-level result.
5
Correct the rule, input, permission or budget causing the problem. Recheck whether an external action already happened before retrying.

​
Example

A delivery question repeatedly goes to sales: inspect its classification and rule conditions, then correct the topic or route.

​
Things to know

  • Classification describes a conversation; routing decides where it goes.
  • An uncertain external result is different from a confirmed failure. Verify it in the connected service before repeating the action.

​
What happens next

Your team can explain automated assignments and identify which configuration needs attention.

​
If something goes wrong

  • Classification missing: check active policy, language, opt-in and budget.
  • Repeated failures: pause the affected workflow while fixing its failing step.

Build an automated workflow

Perform repeatable steps when a conversation, message, ticket or timer event occurs.

Set up AI, credits and your own API key

Choose how AI is paid for and which assistance your workspace can use.