The gap between a chatbot and an agent is action. A chatbot answers. An agent answers, then updates a record, books a slot, escalates to a human or triggers a sequence.
Designing an agent starts with the actions it is allowed to take, not the personality it is supposed to have.
Constrain the knowledge source. An agent grounded in your published, maintained content is easier to trust and easier to correct than one improvising from general knowledge.
Then instrument it. If you cannot see what the agent was asked and what it did, you cannot improve it.