What is your goal?
Does anyone have any ideas on how to prevent Tool/Scenario failure when the ai agent is interrupted through a chat while in the middle of processing a request/using a particular tool or scenario.
What is the problem & what have you tried?
When I test an Ai agent by interrupting it with something like “hurry up”, or whatever a user might say, while it is still going through the process of running various tools and scenarios that it needs to accomplish the task, I get errors (mainly because the agent has already deleted or done something and then (after interruption) attempts to use the tool again.
Error messages or input/output bundles
The interruption causes Tools to return errors because agent either reattempts to use them (ex: reattempts to delete a record that may have already been deleted before agent was interrupted) or the agent stops going through the entire workflow because it gets confused.
Screenshots (scenario flow, module settings, errors)