2018-08-18 at 12:59 AM UTC
I'm writing a shell. Something like you'd get after popping a box with some RCE with MSF. What are some good features to have on/in a shell?
So far i am encrypting the traffic from the client to the handler with AES-256. I have your standard operations to invoke a regular OS shell to be controlled remotely of course. It's multi-platform and i check for the existence of debugger procs and things like tcpview wireshark and stuff like that. If found i have the client self-destruct, especially if it turns out we're in a VM. I was thinking for the lulz if we are not in a VM i'd just brick the device to spite the owner. I got my crypto routines i could expand them to just encrypt everything it can, or do something more sophisticated.
What i also have is 10 stagers, that ship with the client, the purpose of which is to download and execute other kinds of malware. Oh and before i forget, since the shell is multi-platform, i wrote a function that enumerates the system of course. Since i need to know what to execute where. I can expand on that to collect PII or whatever.
But for now, just in your estimation what are some cool/good features for a shell to have in an OffSec type of scenario.
The following users say it would be alright if the author of this
post didn't die in a fire!
2018-08-19 at 12:33 AM UTC
I'll be using something unique to facilitate C2, it involves tunneling to a third party's network and having my handler and client basically connect there. The handler will of course proxychain to the third party but the fun part is that the client sets up all the connections so it looks like legitimate traffic coming from the victim box. Like i said i proxychain to the thir party for opsec purposes. This way i don't need to maintain a bunch of jump boxes or VPS'. Or bother with Tor other than from the handler's end.
Inquire for details via PM, i am not open sourcing this just yet.
The red ones are the best.
2018-08-20 at 8:26 PM UTC
-SpectraL
coward
[the spuriously bluish-lilac bushman]
I did all this 20 years ago. In mine, you could run applications remotely on the mule, such as portscanner and registry editor, and then invisibly "pipe" the outputs elsewhere for pickup. I also had a "paint live desktop", where the mules's desktop could be replicated in real-time anywhere else on the Internet, and also allowed taking remote control of the remote desktop's mouse and clicks. I also had a self-destruct on mine.
2018-08-20 at 9:44 PM UTC
I actually made a small one for Android. It was nothing but an open socket that allowed you to run system commands.
2018-08-20 at 10:02 PM UTC
Cootehill
African Astronaut
[my unsymmetrically blurry oregano]
Do you mean like Bash or tcsh?
If you can make your shell look like Bash, act like Bash, and maybe perhaps even be Bash, that'd be just great.