Can a Rovo agent be restricted from generating its own follow-up questions?

[Jan]

I am working with a Rovo agent and would like to understand whether its conversational behavior can be constrained.

My use case requires that the agent only asks questions that I explicitly define (for example, from a predefined list or scripted flow). I want to prevent the agent from generating additional follow-up or clarifying questions on its own during the chat.

Specifically:

• Is it possible to disable or limit the agent’s autonomous question generation?
• Can a Rovo agent be configured to ask only predefined questions and otherwise wait for user input?
• Are there recommended patterns (e.g. prompt structure, guardrails, or configuration options) to enforce this behavior?

Any guidance, best practices, or confirmation of current limitations would be helpful.

[Jorge Cammarota · Accepted solution]

Currently, there isn't a configuration "switch" in Rovo that simply disables all autonomous generation of follow-up questions, but it's possible to get very close to the behavior you described using a combination of:

very strict instructions in the agent (and in each scenario),

designing the conversation flow outside the agent (for example, in a form, app, or automation that calls the agent),

and, when possible, limiting the agent's role to answering, not leading the conversation.

[Jan]

Hi Jorge,

thanks a lot. I will look at my instructions again and try to make them as strict as possible.