Mirrors that are incomplete (perhaps due to robots.txt, java applets etc.) will
reach out to the original site when viewed on a live system. Is there any
capture option that will prevent this? 

Similarly, is there any capture option that will prevent an incomplete mirror
from reaching out to any external sites when viewed on a live system? I am
familiar with the "no external pages" and "get-non HTML related files related
to a link" options. 

Ideally, I would like to be able to perform captures and be 100% confident
that they will not reach out to any external sources when viewed on a live
system.