An agent action calls a Flow that requires an Appointment Id and a new date. The pet owner has said "move Bella's Thursday appointment to next Tuesday." What does the reasoning engine do about the required inputs?
Select an answer to reveal the explanation.
Short Explanation
You can have the agent use the appointment details it understands from your request, then ask you for any required information it still lacks. Input resolution happens before the Flow runs, so the agent does not need to send your whole sentence as one input or rely on empty inputs and defaults.
Full Explanation
Before invoking the action, the reasoning engine resolves its required inputs: it uses information available from the conversation, context, and variables, then asks you for anything that remains unresolved. Here, it can identify Bella and the requested new date, but those details do not necessarily provide the required Appointment Id. Clear input descriptions help the engine understand what each parameter means and what to ask you for. Requiring a variable to supply both inputs is not necessary; variables can provide known values, but they are not a precondition for planning the action. Passing the entire utterance as one text input is also unsuitable because it sidesteps the action’s distinct, typed inputs and leaves the Flow to interpret the request. Invoking the action with empty inputs and relying on defaults may fail validation or update the wrong record. Exam caveat: information that sounds specific in natural language, such as a pet’s name and a weekday, may still not uniquely identify a record. Before the update runs, check that the Appointment Id matches the intended appointment; if Bella has multiple upcoming appointments, confirm which one the owner means.