When we read a conversation via the API/connector, the message body comes back truncated to roughly 200 characters.
That truncation means the assistant sometimes drafts without the full context — e.g. it asked a customer for a photo when, further down the same email, she'd already said she didn't have one.
Is there a way to retrieve the complete message body through the API/connector (a full-text field, or a parameter to disable truncation)? That single change would make automated drafting much more reliable for us.