I was wondering since no one in #kde-devel can answer it. But when you go to implement Kross into an application. Are Kross scripts like in python, ruby, etc. Do they only execute one time or can you make the scripts run all through the application until you close the application or send a stop signal to Kross to stop them?

I noticed that in the techbase tutorials that you can, well atleast in the examples have multiple interpreters. Are all of them activated upon starting an application instance of Kross? Or what?

The purpose seems to be that they run all the time, or gracefully die a sudden death when somebody tries to retrigger the Kross::Action. KJSEmbed does really funny things and thus hinders further checks here.
To debug them is a real pain in the neck even though you rund them as described in tutorials. If only somebody from the Kross team could clarify this mess or praytell us that they ditched it a long time ago...