Modifying the confidence of intents depending on metadata sent from endpoint

Is it possible to isolate certain intents available to the model and modify its confidence results depending metadata sent from the endpoint.

The use case would be- depending on the location of the user on the site and the contents on the screen, some intents would be more likely to occur.

From the endpoint an example could be:

{ “text”:” “how do I do this?” “location”:”help,core” }

In this case the intents with attached metadata help would be prioritised by some factor over the intents with core. Maybe this kind of behaviour is already implemented? If you have any questions please let me know.