This is a work in progress but it works perfectly fine in its current state. My methodology focuses on creating useful variables for storing stuff, organizing actual actions into "functions," and organizing actual commands into (you guessed it) "commands." The commands reference the functions which reference the variables. I'm no professional programmer, but this works for me. The number of derived commands can definitely be cut down if less spoken command flexibility is allowed.
Enjoy!