Friday 29 April 2016

Walkthrough Droopy v.02

Walkthrough Droopy v.02 :

In this post i shall outline the steps i followed to get root on Droopy hosted at vulnhub . https://www.vulnhub.com/entry/droopy-v02,143/.

From this post onwards i shall try to enlist all the useful links i used to compromise the box for posterity.

Detailed Steps to root :

A nmap scan of the box reveals that only port 80 is open and there seems to be a Drupal 7 website hosted.  After running nikto and dirbuster i decided to run a tool to enumerate CMS's. I know a good one exists for wordpress but surpisingly i could not find much for Drupal. I did come across a tool called CMSMap which according to this github page enumerates wordpress, drupal and joomla . I decided to give this a try to see any information it could reveal. Also i really wanted to be able to learn about Drupal enumeration tools and techniques.

Since the tool is not already built in Kali i decided to follow the github instructions and build it on my box.



The tool was quite good at enumerating the Drupal website and immediately tells us that the site is outdated and suffers from a known vulnerability "Drupal Vulnerable to SA-CORE-2014-005".

Seems like there exists a possibility of launching SQL injection attacks against the box. I decided to use an existing exploit against the box. The exploit https://www.exploit-db.com/exploits/34984/ helps to change the admin credentials on the backend site.



The exploit worked and now we have changed the admin credentials on the box. We now login to the box with credentials "admin:hello".

Once we are able to login to the box we are able to create additional posts/pages. Checking under the "Modules" tab i see that there is a plugin called "phpfilter" which can be enabled to allow php content to be executed on the webpage. This opens the possibility of executing some php reverse shell code by creating a webpage and executing the same via the webserver. Hence we try this approach. I use standard reverse shell from pentest monkey.




Once we get a reverse shell we see that the box is a 64bit ubuntu 14.04 box which is known to be vulnerable to local priv escaltion exploit https://www.exploit-db.com/exploits/37292/ We compile and run the exploit to get root on the box .


Also for completeness I must admit I could have also used a Metasploit module to get a shell but i wanted to avoid using metasploit as much as possible :0

Final Verdict :

Overall it was relatively easy box to pawn but fun nevertheless. Thanks for the creator to take time to create one . Keep up the good work "knightmare" .

Useful Links :

https://github.com/Dionach/CMSmap
https://www.drupal.org/SA-CORE-2014-005
https://www.exploit-db.com/exploits/37292/
https://www.rapid7.com/db/modules/exploit/multi/http/drupal_drupageddon

Wednesday 13 April 2016

Z-wave protocol analysis using Ez-Wave

Background: Recently I was able to get my hands on a couple of IoT devices talking the propriety Z-wave protocol. Specifically I had 2 devices -

i) Z-Wave 3 in 1 PIR Motion sensor  which can detect movement, temperature and luminance.
ii) Z-Wave smart Energy Plug which can give information about the energy cosumption of the device connected to it. Also it can switch the device ON/OFF depending on the requests sent to it by a Z-Wave controller.

These devices are already deployed in many smart homes. Hence its important to analyze the security capabilities of these devices. This got me rolling and here is post about how to sniff Z-wave packets using SDR. It seems that the state of the art in pentesting any wireless comunnication protocol is by using these SDRs.

A detailed security assessment of Z-Wave protocol and Z-Wave enabled devices was presented at ShamooCon 2016 Breaking Bulbs Briskly by Bogus Broadcasts https://www.youtube.com/watch?v=IgquSEhAGvA by Joseph Hall and Ben Ramsey which details the how many Z-Wave device manufacturers disregard security and dont even encrypt the data exchanged between the Z-wave device and the Z-wave controller. This opens up possibilities of a range of attacks that can be launched on the Z-wave devices. They also release a open source tool to play around and sniff these packets called EZ-Wave . In this post i shall try to sniff some of the traffic generated by my Z-Wave devices.

EZ-wave Installation:

Install EZ-wave dependencies:

The instructions for installation are detailed on the github site @ https://github.com/AFITWiSec/EZ-Wave. However there are a number of software bundles that need to be installed especially the Gnu-radio which has a lot of dependencies. Hence instead of installing all these packages myself i used a linux distribution Pentoo Linux. (http://www.pentoo.ch/about). The advantage of using the Pentoo linux is that all the software requited for EZ-Wave installation such as GNU Radio, OsmocomSDR, HackRF host software, Wireshark etc  are all  pre-installed in the distro saving the time to install them manually.

The simplest way is to download the pentoo linux package from the downloads section of the site and then burn the image into a live USB stick. One important thing to note here is that Michael Ossmann (maker of HackRF) highly recommends in his "Introduction to SDR " tutorial (https://greatscottgadgets.com/sdr/) to not connect the HackRF to any virtual machine for performance reasons but to use it directly on the base OS.  Also on a side note if you are a newbie its highly recommended to go through at least his first tutorial where he introduces the SDR concepts  (https://greatscottgadgets.com/sdr/1/) and also the tutorial on HackRF usage. (https://greatscottgadgets.com/sdr/5/) .

We now boot from the live USB. (In case   you get an error like boot device not found press "Tab" on the keyboard and select pentoo5 OS or type it and press enter). The Pentoo linux should boot correctly. Launch an GUI using the "startx" command. In case there is no IP address allocated to the box just do a "dhcpcd eth0" to request an IP address.

Now to test that we have the HackRF software working correctly plugin the hackrf devices into the USB port and run the command "hackrf_info" . I use 2 Hackrfs since they are half-duplex and i wish to receive and transmit at the same time.


Install Scapy-Radio 

We can see that both our devices are correctly detected. Next we install the EZ-wave tool itself. We run the setup.sh script which clones the scapy framework
and then we install it.



I noticed that if you use pentoo linux then you dont need to worry about making changes to the gnu-radio config file.

Install Wireshark

We downloaded the source package for wireshark version 1.12.10 from https://www.wireshark.org/download/src/.  Copy the wireshark dissectors files to the wireshark-1.12.10/epan/dissectors. Wireshark by default tries to use Qt version5 for building the wireshark UI. This gave me errors so i choose to make my wireshark ui with gtk3. Hence its important to tell wireshark to use gtk3 library instead of qt5 libraries at the config stage. Hence our steps to build wireshark were



EZ-Wave Usage:

In order to sniff packets we start gnu-radio companion  and provide it as input the Zwave radio configuration file.

Select the Zwave.grc file to open in the gnuradio-companion

One important thing to take note of in the .grc file is the variable central frequency i.e center_freq . The European standard Z-wave devices ( which I have ) talk on 868.42 Mhz ( instead of the default central frequency hardcoded in the .grc file which is for the US Z-wave devices) and hence this variable needs to be changed to reflect the frequency of your device. To change the variable double click it and then key in the new value.

gnuradio-companion main window
After changing the .grc file we need to recompile the flow graph. Finally we can run it by pressing the play button on the top menu. (Also make sure that the python modules are in the python path environment variable else there may be some errors)

The HackRFs should now be able to pick up Z-wave packets in the vicinity. I forced the Z-wave device to send some packets by pressing the button on the device body. All the captured Z-wave packets are sent to localhost:52002 and can be seen using the wireshark and the new dissector. You can apply some filters like !icmp to remove the interrogation commands sent to the Z-wave devices.

Wireshark Z-wave packet capture

We can see the homid and nodeid of the captured Z-wave packet. The data exchanged between the Z-wave controller and the Z-wave device is also not encrypted.


Sunday 21 February 2016

Walkthrough LampSecurity Version 7

CTF series LAMP SECURITY 7 : I have been spending a lot of time trying to improve my hacking skills in order to prepare for my OFFSEC certification challenge. I have been struggling to improve my average time to compromise  a box ( remember i need to do 5 in 24 hours to clear the offsec challenge) so any easy boxes i find along the way are welcome :P . They kinda increase my confidence. Lampsecurity7 seems like one such box. I am glad i pawned it and that too in optimal time (imho since the time can greatly vary depending upon the skill of the person). Thanks to the author for helping me prepare for the exam :-)

Detailed Steps to get root :

An nmap scan of the box reveals a number of open services.



We see a webapplication running on the box. While playing around with the webapplication we notice the /newsletter URL which by adding a ' to the URL we a detailed error message printed out which leads us to believe that the box suffers from sql injection.



We capture the traffic using burp and run sqlmap to see if we can find any interesting databases.





Sqlmap is able to dump a few databases for us namely i) information_schema
ii) mysql iii) roundcube  & iv) website

We try to dump the contents of these databases using the sqlmap -D parameter 



Sqlmap is able to dump various tables in the website database. We keep dumping the contents of these various tables till we see that the `users` table dumps a list of possible database users and their credentials (sqlmap helps us crack these as well . its so amazing and powerful !) . Luckily we see that the passwords are not salted.

We know that the box is running a ssh service so we try to login to the box hoping that at least some user would have reused his ssh credentials.

We ssh into the box and seems like julia is in the sudo users list and we can immediately escalate our priv to root :-) Simple and easy




Sunday 7 February 2016

Walkthrough TopHatSec Freshly

Mission `Freshly` VM : I downloaded the virtual machine hosted at https://www.vulnhub.com/entry/tophatsec-freshly,118/  . There are a couple of challenges from TopHatSec and I would like to try them both. Well for starters here is `Freshly` :-)

Detailed steps for getting root : 

A nmap scan of the box reveals that ports 80,443 and 8080 are publically accessible  on the box.


We quickly run both nikto and dirbuster against the webserver. We see that the VM is running a login application on port 80 while on the other ports it is  running a wordpress blog.


Login prompt 
Wordpress blog
We also see an installation of  phpmyadmin on the webserver. We try the default credentials but dont succeed. Since there is a phpmydamin running on the box we guess that there may be mysql server running at the backend as well. Hence we try mysql authentication bypass by using a wronguser' or 1=1 LIMIT 1;# as username and junk as password. However we are not successful. Inspite of this its worth the effort to run sqlmap against the server to see if it can pick up some injection points either in the username or password fields.



Sqlmap tells us that the `user` parameter of the login form is sucesptibel to a mysql injection vulnerabilty. It readily exploits this vulnerability for us to give us a complete dump of the database when we use the --dump sqlmap option.



Excellent! now we can see that there are a total of 7 databases in the database. We must check out the interesting ones especially login, phpmyadmin, users and wordpress8080.



It seems that the table dump for the wordpress8080 database contains the admin password for the wordpress blog.  We try to login to the wordpress blog via the credentials and we are successful!


Once we have admin credentails on the wordpress getting a shell on the box is extremely simple. By default wordpress will not allow any php content to be run from the wordpress articles/posts. However since we have admin privs on the wordpress we can install a plugin called Exec-php to allow php code to run in the wordpress pages.

We install the plugin to our wordpress installation and then follow the instructions at https://wordpress.org/plugins/exec-php/installation/ to run php code in the article body. Once the simple php hello world example works we use the php-reverse-shell script available at http://pentestmonkey.net/tools/web-shells/php-reverse-shell . We set up a netcat listener and sure enough we get our shell :-)



There we have our flag. A fun challenge indeed :-) 

Wednesday 27 January 2016

Walkthrough Acid server



Mission : Get the root flag on `Acid` which is a deliberately vulnerable virtual machine hosted at https://www.vulnhub.com/. The virtual machine can be downloaded at https://www.vulnhub.com/entry/acid-server,125/. I quickly loaded up the virtual machine into my VMWare Player and i was good to go !

Detailed Steps for getting root:

A nmap scan of the box reveals only one service running on a non default port. It seems that there is only this http service running on the box.



Upon accessing the http service on the port 33447 we are presented with a "Welcome to the world of acid" webpage. 


Acid Page


Our first instinct it to always view the source code of the page to see if there is anything interesting and sure enough we are lucky as in this case we find a hidden comment - "643239334c6d70775a773d3d". It seems that its in hex . We convert hex to ascii to get "d293LmpwZw==" . The == sign at the end tells us that its base64 encoded. We are quickly able to decode the string as wow.jpg. ! Lets see if we can use this image later.

We also see that the page title is "/Challenge" which tells us that such a folder may exist on the webserver. When we visit the /Challenge folder on the webserver we see a login page prompting for an email address and password. We cant do much now so we use dirbuster against the webserver. Our dirbuster tells us there are some interesting files and folders on the webserver. 


Dirbusting /Challenge directory

We also interact with the login page via the burp suite to understand better the way the form interacts with the backend webserver. We can see that the form converts the password into a hash using a api from the includes javascript files namely sha512.js and forms.js. We try to look into these files and see that these files are infact part of a php login system called "phpSecureLogin" https://github.com/peredurabefrog/phpSecureLogin. It seems that the project has been abandoned but we find that the github page (https://github.com/peredurabefrog/phpSecureLogin) lists default credentials (email : test@example.com Password: 6ZaxN2Vzm9NUJT2y ) built into the login system. We use it  and we are successfully able to bypass the login screen ! 


Login form showing the includes javascript files

Once we login we have a look at all the files that are listed by the dirbuster. When we view the cake.php we see that the page title changes to /Magic_Box hinting that there may be something with that name on the webserver. 



It seems that we dont have the permission to view the  /Challenge/Magic_Box directory but we run a dirbuster against it to see if it reveals anything. 


Dirbusting /Magic_Box directory

We are able to find a bunch on interesting files such as low.php, tails.php , command.php etc. It seems that command.php is a console that allows us to ping another IP. We suspect a command injection vulnerability here and hence besides providing a IP address we provide another command such as "id" with the assumption that the backend code will not sanitize the input and not limit execution to a single IP address command.


Command injection

Sure enough we can see that our code is being executed.



Once we can execute code on the server we attempt to get a reverse shell. We reference the reverse shell cheat sheet http://pentestmonkey.net/cheat-sheet/shells/reverse-shell-cheat-sheet and select the perl reverse shell after checking that nc though installed does not support the -e option. To be able to use the perl reverse shell we must url enocde it and also since the length of the IP address field is set to max 200 we need to intercept the request in burp before forwarding it to the webserver. 



Burp intercept


Sure enough we get our low priv shell !


Privilege Escalation : 
Escalating privileges on the box takes quite a long time. Once we have a low shell we peep into all the file present in the Challenge and Magic box directories. We see that the box is running a MySQL service and the file psl-config.php contains the database password. 



In the database we check the secure_login database to find a few users and their password hashes and password salts. We are able to crack the hashes ( saman & Vivek) but since the passwords are salted we wont be able to crack it. At this point we give up on the database. We slowly follow the priv escalation guide https://blog.g0tmi1k.com/2011/08/basic-linux-privilege-escalation/ till we finally see an interesting file on the box via enumerating all the files for the user acid and user Vivek on the box. 



The pcap file is interesting and we download it and open it in wireshark. We guess that we should look into the TCP communication captured on the network interface and hence set the wireshark filter to the same. Then we follow the TCP stream to see the data exchanged in the session. And sure enough we get our next clue.



Looking at the message we guess the user saman password. Also always check if the user is in the sudoers list. This gives us root !









Wednesday 13 January 2016

Walkthrough Vulnix

Mission : Get the root flag on `Pipe` which is a deliberately vulnerable virtual machine hosted at https://www.vulnhub.com/. The virtual machine can be downloaded at https://www.vulnhub.com/entry/hacklab-vulnix,48/. I quickly loaded up the virtual machine into my VMWare Player and i was good to go !

Detailed Steps for getting root:

A nmap scan of the box reveals a number of services running on the box such as ssh, smtp, finger, pop3, imap , rlogin , rexec, rshell and nfs. 


Our attention is drawn to a number of services which we would typically not see exposed such as finger, rlogin services and nfs.NFS protocol allows a user on a client system to access a folder on the network as if it were present locally. Poorly configured NFS services are known be exploitable.[ If we remember there is a exercise in Metaspoitable that deals excursively on attacking the NFS protocol for getting root]. Hence we go for this service and enumerate if we can mount a share locally. 








We see the the /home/vulnix directory  can be mounted and we mount it on our attacking box. However we are unable to see the contents of the directory as we keep getting a permission error. We get this error because we are root on our attacking machine and we are trying to access a non root owned directory on the target system. We guess that this is because the NFS /etc/exports file (https://www.centos.org/docs/5/html/Deployment_Guide-en-US/s1-nfs-server-config-exports.html may have the root_squash permission turned on which squashes the permissions of the root client to the lowest user to prevent unauthorized alteration by a client on the nfs shared directory. In order to view the contents in the mounted directory we need to acess the nfs drive as a user with the same uid and guid as the vulnix user on the target box. Since we dont know the uid and guid of the vulnix user we do some further enumeration on the box. 

We see that the smtp service is running of the box. There exisits a metasploit auxilliary module that helps us enumerate all the valid users accounts on the smtp server. We use this module as follows 



The module has identified a number of user accounts on the box. The `user` account catches our attention. We guess that accounts maye be reused and use the finger service to determine if its possible to login to this box via the `user` account. 



Clearly an account called `user` can be used to login to the box. With no other information available we attempt to bruteforce the ssh service with this user account. After a few minutes we are able to bruteforce successfully ! ! The credentials for this account are user:letmein . Now we can login to the box as `user`. 



Once we login we see another user called vulnix. We take note of the uid and giud of this user (2008 in both case). We shall attempt to create a user with the same uid and guid on our client box. Hopefully this will solve out permissions problem and we will be able to elevate our priv to vulnix . 



Now on the attacking box we create a testuser with the same uid and guid .


Now after we mount the /home/vulnix directory we no  longer see the permission problem !

Saturday 9 January 2016

Walkthrough SecOS: 1

Mission : Get the root flag on `Pipe` which is a deliberately vulnerable virtual machine hosted at https://www.vulnhub.com/. The virtual machine can be downloaded at https://www.vulnhub.com/entry/secos-1,88/. I quickly loaded up the virtual machine into my VMWare Player and i was good to go !

Detailed Steps for getting root:

I ran a quick nmap scan to identify the Ip address of the Pipe VM. I use `host-only` networking for both my Pipe Vm and Kali box.


We identify the IP address of the new VM as 192.168.28.138. 

A nmap scan of the box shows that the box is running a http service on port 8081 and a ssh service on 22.



We run nikto againts the webserver but dont find anything of much interest. We then dirbust the webserver which reveals a bunch of interesting directories that we should look into.



It seems that the vulnerable application allows us to create a user account and have a look at the various pages on the website. We see that among other things there seems to be 3 users on the box one of whom is the administrator. Also there is a way to send messages to the admin. 



We also see an About page that tells us that the site has been developed using Node.js and MangoDB. At this point we try the an authentication bypass seen in Node.js and MangoDB systems. A good discussion of it is here  : Hacking Node.js and MangoDB http://blog.websecurify.com/2014/08/hacking-nodejs-and-mongodb.html: . It seems that the box is not vulnerable to this bypass. We begin to observer the webapp more closely and find that the login page source code has comment which is a hint. 




We have a look at the /hint page and its source code which reveals that we should launch a CSRF kinda attack on the administrator. 




The hint page tells us that the administrator will visit any site running on 127.0.0.1 if he is asked to do so. The hint is clear that we should feed a URL to the administrator to see. This can be done via the messaging feature of the webapp which allows us the send messages to the administrator. Relooking at the webapp we see that there is a feature to change the password. We decicde to trick the admin into changing his password to our specified one. For this we create a HTML page with hidden inputs and POST the data to the /change-password . 



We know ask the admin to view this page by sending him a message . 


We hope that when the admin visits this page his password will get reset and we can login. Well after a few minutes we try logging in as admin and we see that our little trick has worked ! 

We login to the admins message board and see that he has a message from a user 'pirate` who claims to have cracked his password.



We try ssh'ing into the box with these credentials and we are immediately logged in  ! 

Once we login we see that there a bunch of interesting files in the users home directory. 



If we do a quick directory listing 2 things stand out. i) There is an internalServer running as root on the box ii) there is a additional ping.ejs file which is not normally visible via the public facing web interface. 

We look at the contents of both the internerlServer and ping.ejs files.


It seems that besides the public facing webapp there is an additonal server running on the box on port 9000. This server provides a Ping capability for the admin to maybe ping other systems. When we look at the code for the ping service we see that the exec api is called with the user provided IP address without any input checks. We can inject code here as the application does not sanitize the input data. 

We can launch the command injection both via the web browser or the command line. Let see both approaches . 

a) Command line : We can use wget or curl to post data to the internal server and inject our code in the post request. 



Here we send a post request with a ip address =1 and inject our 'id' command along with the request .

Similarly we can use curl to post data as well 



b) Web browser : In order to be able to see the internal webapp on our attacking machine we use a ssh reverse port forwarding technique using the ssh -L parameter. Via this method we make the port 9000 of our victim machine available on our kali machine at port 4545. 



Now we can access the port 9000 of the inernal web application