The add-on works well for eternal chats, because it doesn't know when to end the conversation, which increases the cost of the API (OpenIA).
Even if you send other messages via WhatsHAM or use another flow, the conversation with the AI ​​doesn't end.
The same thing happens when you start another AI flow, the chat is eternal, it needs to be updated so that the conversation ends when it meets some requirement.
The support is fast, it answers via WhatsApp, but it hasn't solved this issue yet.