Create a variety of accessibility API targets as model sources

[C-Loftus]

By using the accessibility APIs within Talon you can essentially retrieve any part of a document without needing to actually select the text and copy it.

Some useful scopes could be the entire document, just the editable text field, or the current element where the focus is present. Some of these may be slightly different depending on application

This would be unsupported on Linux since there are no bindings to the Linux accessibility APIs.