Well, to do it efficiently, before you do any of that you first need the genome sequence for the ball python so you can determine where to best put the GFP tag. And to the best of my knowledge no one has gone out and sequenced the whole ball genome yet (anyone got a spare million bucks and the willingness to lend a few normal BPs for blood draws?)
Once you have that the process should actually be pretty "straight forward" in terms of already existing techniques.