On remote execution

[jasonwitty]

At least for the time being, I have decided to not allow any remote execution through the socktop agent. Later as the project becomes more mature I may change my decision on this if a the user supplies a token and https connections are enabled. But for the time being I want the agent to remain "read only" to avoid any potential security issues.

I will be implementing the ability to kill a process on your local machine. This is to allow you to use socktop as a single top solution for your local machine as well as remote machines. This kill will not pass through the agent, so you will only be allowed this feature on local for the time being.