Is there any keywords or special symbols supported by the avatar?

Hi folks,

I'd like to use custom knowledge with avatar. Navigation links are part of my custom knowledge, e.g. I can tell my avatar "navigate to documents page" and avatar will reply "/documents|Documents section represents..." (this answer will be taken from my knowledge).

"/documents|" will be captured on client side and JavaScript code will do the navigation while avatar speaks. This part of the answer should not be pronounced by avatar, avatar should say everything after |.

What would you recommend to solve this issue?