Selenium powered Python script to automate searching the web for vulnerable applications.
DorkNet can take a single dork or a list of dorks as arguments. After the proper command line arguments have been passed, the script will use Selenium and Geckodriver to find the results we want and save them to a textfile for further processing with SQLmap or similar utilities.
git clone https://github.com/NullArray/DorkNet.git
cd DorkNet
python dorknet.py
The options for the program are as follows.
-h, --help show this help message and exit
-d DORK, --dork DORK specify the dork you wish to use
-l LIST, --list LIST specify path to list with dorks
-v, --verbose toggle verbosity
Some examples for clarity.
DorkNet.py -h
DorkNet.py -d inurl:show.php?id= -v
DorkNet.py -l /path/to/list.txt --verbose
Proxifying
I have included the ability to proxy the connection of the web driver if desired. Simply provide the proxy IP and PORT when the dialog comes up and the search engine will be accessed via the proxy settings you have provided.
You will need the Mozilla Geckodriver for this to work. You can install it manually. However i've added a shell script to automate the process if you'd prefer.
To use it, navigate to the DorkNet directory and make the shell script executable like so.
cd DorkNet
chmod +x gecko-setup.sh
# Execute the shell script with the below command.
sudo ./gecko-setup.sh
The shell script has an option to automatically install the rest of the dependencies as well, via the requirements file, by invoking the following commands.
sudo -H pip install -r requirements.txt
I've added some logic that lets the user run Geckodriver in Headless Mode, that is to say, without the traditional UI. This is useful if you have a particularly long list of dorks you'd like to work with. In it's current implementation, the function that is responsible for proxyfying our connection to Google interferes with the operations required to set the Geckodriver options to enable headless mode. Therefore, running DorkNet with the --nogui flag and a proxy enabled will not work as it is supposed to.
However, users that would like to run the program in headless mode anyway, can. I've committed all the relevant code to the repo. For the tme being it's just been commented out. Remove the comments and it should work without issue. Check out the commit historry here to see the affected lines.
Now if you really want to proxy your connection while --nogui is set, you can start DorkNet with ProxyChains like so;
proxychains python dorknet.py -l /path/to/dorks.list --nogui --verbose
If you don't have ProxyChains simply follow the link above or use your package manager to install it. Thank you.
DorkNet is featured in the BlackArch Linux PenTesting Distro under WebApp Tools & Automation. As such it comes pre-installed with the distro. Refer to the relevant PKGBUILD file in it's respective repo for details.
By using Selenium and Geckodriver, DorkNet is effective at emulating a regular browser. In this manner the program is able to avoid captchas most of the time. However on limited occasions, Google throws one regardless. The same sometimes happens when manually searching for strings that look like a dork. Should you encounter one, you can just fill out the captcha in the Geckodriver and DorkNet will continue it's normal operation.