README.md

httphish

Quickly clone a website and launch an HTTP server to phish information with httphish.py ?

Only one Python 3 script with no dependencies! Simply run the script, it will automatically download a web page, host it on a local HTTP server and save all POST data sent by visitors (such as login credentials).

Currently, it only works on simple pages with <form> logins. It might work on some dynamically loaded pages if they aren't too complex. Please see at the bottom of this README for examples of public websites that work and don't work. Also, the wget command is currently required to download websites, so this feature only works on Linux. I will probably eventually add OS detection and use Invoke-WebRequest from PowerShell on Windows, but for now, please download websites manually. (Pull requests are welcome!)

How to use

Clone this git repository to download the necessary files and run the script:

git clone https://github.com/thom-s/httphish
cd httphish
sudo python3 httphish.py

It will then ask you for the following information :

• Whether you want to download the webpage with wget or if you have manually saved it to the /web folder.
  • If you use wget it will also ask you :
    • The full URL to download (ex: http://www.github.com/login)
    • Whether to use the default user agent for wget or enter a custom one. (You can see the default one in the code)
  • If you want to manually download it, simply create a folder named web next to the script and save index.html in it.
• The IP/domain to redirect all GET/POST requests to. If any files cannot be served statically, it will HTTP forward the request there. (ex: www.github.com)

The <form> tags in index.html will then be modified to redirect requests to localhost.

You will then be prompted to press Enter to launch the HTTP server.

Browse to your own IP address (or localhost) and you will see a cloned version of the website.

Received POST and GET requests will be displayed, and POST requests coming from forms in index.html will be saved to post.txt. After running httphish.py, simply run the following command to see them : cat post.txt

When you are done, press CTRL+C to close the HTTP server and end the script.

Before running it again, simply run cleanup.py to delete the /web folder and the post.txt file :

sudo python3 cleanup.py

Use case

This script serves as a tool for pentesters to phish credentials for certain websites. Such a server is simple enough to be hosted from any device capable of running python. For example, you could have a Raspberry Pi on the local network, controlled remotely by you, which copies an internal company website and hosts the clone. Internal corporate websites are generally simple enough to be cloned; this means phishing attemps by e-mail links or even DNS/ARP poisoning can generate system credentials, maybe even AD credentials.

What system administrators can take away from this, is how technically simple convincing phishing attacks can be. Thankfully, a combination of security measures should be able to stop these attacks. Aside from the usual e-mail filters and dns/arp poisoning preventions, system administrators can train users to recognize phishing e-mails and phishing websites. Common network security measures should be able to mitigate this.

Troubleshooting

• Some websites that do not work when you automatically download them might work if you manually save them.

• Some websites won't respond to requests directed to their IP, so try entering the domain instead (or vice-versa).

• Some lazy-loaded content simply doesn't work.

• If a website doesn't work, use inspect element and look under the network tab. The issue is probably some dynamic requests being broken because the site is too complex.

  • In some cases, this can be fixed by changing the IP/domain to redirect GET/POST requests to.
  • In most cases, you would have to manually modify the files and choose to not automatically download the file.

Website examples

Working websites

Working websites will generally have very simple login forms and not much dynamically loaded content. Here are some I tested.

• http://www.github.com/login
• http://www.linkedin.com/
• http://www.facebook.com
• http://old.reddit.com/login

Partially working websites

These sites will work, but some content might not get loaded.

• http://www.stackoverflow.com/users/login

Broken websites

For most broken websites, dynamically loaded content will be the issue. Here's some websites I found did not work.

• http://www.instagram.com/accounts/login
• http://www.messenger.com/login
• http://www.reddit.com/login