Create the "convenience"-APIs for a "LocalSystemAgent" to guide people into transparency when doing something dangerous like local system actions

[frankhaugen]

To have an application that can use AI to control the local machine, (run scripts with ProcessInfo, or even lower level system interopt calls), would be fantastic for some uses, HOWEVER this is dangerous because if I make an agent/plugin/connector (or combo) that has the ability to run a command like shutdown -s -f -t 0 on a server... It gives me goosebumps just thinking about it...

So, I have an idea: I would like to see the library supplied some interfaces/base-classes and helpers, that make this sort of thing easy and is well-promoted so people will use it, because, if the lazy and simple way is built-in, SK can ensure there are logging and some mechanisms to prevent "forbidden action" from being run.

Another part will be that these things could get logging and telemetry into it so it would be more "transparent"