HOON is a python shellcode creator. It does not just create shellcode, it can parse AT&T's syntax assembler and create the shellcode from that.
HOON is a product of not having a decent PUBLIC shellcode creator and it works pretty well too. Some things to note though, it does not have a huge instruction/opcode database and this will be fixed in future releases. It also does not handle jump/call's very well or addressing (movl (%eax),%esi for example) these are minor issues. Most definitely have a look at the tokening system - if someone can port this to Intel syntax, it would be great.
To add more entries, go to the init (not __init__) method and use self.addreg (ignore the name ;)) with the instruction/format of instruction/opcode. Opcode will need a base (commonly \xc0) if it's all register instruction.
single instruction: self.addreg("int3",self.format["single"],"\xcc")
instruction with format instruction reg,reg self.addreg("movl",self.format["insregreg"],"\xc7\xc0")
* note, Python (not ActivePython) doesn't like values with it's strips() so change that will a replace(self.null,"")
