Well, injecting into http will be useless, but you may certainly try it of course. Each login has coherency count embedded, you will see.
Anyways, nice to see some creativity.
As for I_Learning_I :
One kill? Uhm. There was never (even in first server released 2005) an exploit that enabled you to one hit kill somebody.
Let's leave out weather or not it's worth to buy my bot - you guys already decided that by posting here. You do not need a constant internet connection to protect your program against anything. I can tell you so far, the auth only takes place once. Internally, the licence is validated several times, even while botting, using an X.509 PKI to validate the core code and the licence checking code.
Keep going, I enjoy this thread very much
Swoosh it's not my objective to crack your tool, it's your option to make a paid hack and if someone asks how to crack a tool I give out the basic knowledge about it. As much as I like to help, in this case helping one will harm the other, therefore I will not crack your tool (you can see that I haven't gone in detail about it).
Anyhow if you're doing a local check your program is crackable, it might take 10000 bypasses, but at the end of the day it's possible.
I don't know about X.509 validation, had a quick look at wiki, but I don't know (in detail) how it works.
About that hack... It was real, it was also very private, search away you'll never find it, it has never been posted in any forums/website.
Good luck to the cracker and to the coder.