First Iteration of ChatGPT Generated HTTP/S Beaconing Implant.
All the code was generated with ChatGPT and then hand-fixed as needed to get a minimum viable proof of concept working.
This is version 1, which is entirely Python for now, intended purely as a playground/testbench.
You can find the writeup on this iteration on my blog: www.fullspectrum.dev/chatgpt-assisted-implant-dev-part-1/
This is probably going to be "frozen in time" after a followup blog post, with a new repo for the next "version" in this experiment.
Works only on Linux targets.
- Command Execution
- Beacons every minute (you can change this in code)
- Nothing else
- No encryption
- No authentication
- Literally a proof of concept
- Full of bugs.
Install flask, sqlite, requests - see requirements.txt
-
Run mkdb.py to create the database.
-
Run c2.py to launch the command and control server.
-
Edit bot.py to point to your command and control server.
-
Edit control.py to point to your command and control server.
-
Run bot.py on your target.
-
Run control.py and use the arguments to list agents, list tasks, add tasks for agents, get output from tasks, cancel tasks...
Licence: WTFPL
Bugs? I guess make an issue, I'll probably be interested.