marciopocebon / lil-pwny

Fast, offline auditing of Active Directory passwords using Python.

Lil Pwny

Fast, offline auditing of Active Directory passwords using Python.

About Lil Pwny

Lil Pwny is a Python application to perform an offline audit of NTLM hashes of users' passwords, recovered from Active Directory, against known compromised passwords from Have I Been Pwned. Results will be output in JSON format containing the username, matching hash (can be obfuscated), and how many times the matching password has been seen in HIBP

There are also additional features:

• Ability to provide a list of your own custom passwords to check AD users against. This allows you to check user passwords against passwords relevant to your organisation that you suspect people might be using. These are NTLM hashed, and AD hashes are then compared with this as well as the HIBP hashes.
• Return a list of accounts using the same passwords. Useful for finding users using the same password for their administrative and standard accounts.
• Obfuscate hashes in output, for if you don't want to handle or store live user NTLM hashes.

More information about Lil Pwny can be found on my blog

Resources

This application has been developed to make the most of multiprocessing in Python, with the aim of it working as fast as possible on consumer level hardware.

Because it uses multiprocessing, the more cores you have available, the faster Lil Pwny should run. I have still had very good results with a low number of logical cores:

• Test env of ~8500 AD accounts and HIBP list of 613,584,246 hashes:
  • 6 logical cores - 0:05:57.640813
  • 12 logical cores - 0:04:28.579201

Output

Lil Pwny will output results as JSON format either to stdout or to file:

{"localtime": "2021-00-00 00:00:00,000", "level": "NOTIFY", "source": "Lil Pwny", "match_type": "hibp", "detection_data": {"username": "RICKON.STARK", "hash": "0C02C50B2B08F2979DFDE12EDA472FC1", "matches_in_hibp": "24230577", "obfuscated": "True"}}

This JSON formatted logging can be easily ingested in to a SIEM or other log analysis tool, and can be fed to other scripts or platforms for automated resolution actions.

Installation

Install via pip

pip install lil-pwny

Usage

Lil-pwny will be installed as a global command, use as follows:

usage: lil-pwny [-h] -hibp HIBP [-c CUSTOM] -ad AD_HASHES [-d]
                   [-output {file,stdout}] [-o]

optional arguments:
  -h, --help            show this help message and exit
  -hibp HIBP, --hibp-path HIBP
                        The HIBP .txt file of NTLM hashes
  -c CUSTOM, --custom CUSTOM
                        .txt file containing additional custom passwords to
                        check for
  -ad AD_HASHES, --ad-hashes AD_HASHES
                        The NTLM hashes from of AD users
  -d, --duplicates      Output a list of duplicate password users
  -output {file,stdout}, --output {file,stdout}
                        Where to send results
  -o, --obfuscate       Obfuscate hashes from discovered matches by hashing
                        with a random salt

Example:

lil-pwny -hibp ~/hibp_hashes.txt -ad ~/ad_user_hashes.txt -c ~/custom_passwords.txt -output stdout -do

Getting input files

Step 1: Get an IFM AD database dump

On a domain controller use ntdsutil to generate an IFM dump of your AD domain. Run the following in an elevated PowerShell window:

ntdsutil
activate instance ntds
ifm
create full **output path**

Step 2: Recover NTLM hashes from this output

To recover the NTLM hashes from the AD IFM data, the Powershell module DSInternals is required.

Once installed, use the SYSTEM hive in the IFM data to recover the hashes in the format usernme:hash and save them to the file ad_ntlm_hashes.txt

$bootKey = Get-BootKey -SystemHivePath '.\registry\SYSTEM'
Get-ADDBAccount -All -DBPath '.\Active Directory\ntds.dit' -BootKey $bootKey | Format-Custom -View HashcatNT | Out-File ad_ntlm_hashes.txt -Encoding ASCII

Step 3: Download the latest HIBP hash file

The file can be downloaded from here

The latest version of the hash file contains around 613 million hashes.

Resources

• ntdsutil & IFM