Difference between revisions of "Alternate XSS Syntax"

Jump to: navigation, search
Line 6: Line 6:
[[ASDR Table of Contents]]__TOC__
==Related Security Activities==
==Related Security Activities==

Revision as of 06:31, 7 April 2009

This page was marked to be reviewed for deletion.

#REDIRECT Cross-site_Scripting_(XSS)

Last revision (mm/dd/yy): 04/7/2009

Related Security Activities

Description of Cross-site Scripting Vulnerabilities

See the OWASP article on Cross-site Scripting (XSS) Vulnerabilities.

How to Avoid Cross-site scripting Vulnerabilities

See the OWASP Development Guide article on Phishing.

See the OWASP Development Guide article on Data Validation.

How to Review Code for Cross-site scripting Vulnerabilities

See the OWASP Code Review Guide article on how to Avoid Reviewing code for Cross-site scripting Vulnerabilities.

How to Test for Cross-site scripting Vulnerabilities

See the OWASP Testing Guide article on how to Test for Cross site scripting Vulnerabilities.


Cross-site Scripting (XSS) attacks have to pass two hurdles. First they must get through the application without being filtered, validated, or encoded in a way that prevents them from executing. Second, they much be carefully crafted so that they seamlessly insert their payload into the HTML document so that it will run when loaded by the browser.

Cross Site Scripting is not just <script>alert('XSS');</script>. Because of JavaScript and HTML flexibility and their interpretation by the web browsers, it's possible to achieve the same goal in many different ways. These attacks extend the basic XSS attacks with a number of alternate encodings to help them pass through the application.

In effect we may try to bypass more or less successful input data filtering methods. Some JavaScript and HTML constructions after encoding are correctly interpreted by some browsers, nonetheless it often varies on the web browser version, and others are not.

If we want to use popular <script> tags anyway, we may try to bypass filtering replacing given characters with their equivalents:

From To
<    &lt;
>    &gt;
(    &#40;
)    &#41;
#    &#35;
&    &amp;
"    &quot;

In this case:

 <script>alert('y0u ar3 0wn3d!');</script>

would be replaced with:

&lt;script&gt;alert&#40;'y0u ar3 0wn3d!'&#41;&lt;/script&gt;

In most contexts this encoded string will not execute. However there are some environments, particularly Ajax and XML processing engines, that will automatically decode these encoded characters and allow the attack to execute.

We don't need to do replacement at all, we may get the same effect in many different ways.

Risk Factors


XSS using Script in Attributes

XSS attacks may be conducted without using <script></script> tags. Other tags will do exacly the same thing, e.g.:

<body onload=alert('test1')>

or other attribites like: onmouseover, onerror.


<b onmouseover=alert('Wufff!')>click me!</b>


<img src="http://url.to.file.which/not.exist" onerror=alert(document.cookie);>

XSS using Script Via Encoded URI Schemes

If we need to hide against web application filters we may try to encode string characters, e.g.: a=&#X41 (UTF-8) and use it in IMG tag:

<IMG SRC=j&#X41vascript:alert('test2')>

There are many different UTF-8 encoding notations what give us even more possibilities.

XSS using code encoding

We may encode our script in base64 and place it in META tag. This way we get rid of alert() totally. More information about this method can be found in RFC 2397

<META HTTP-EQUIV="refresh"

These (just a little modified by me) and others examples can be found on http://ha.ckers.org/xss.html, which is a true encyclopedia of the alternate XSS syntax attack.

Related Threat Agents

  • TBD

Related Attacks

Related Vulnerabilities

Related Controls

  • HTML Entity Encoding
  • Use whitelists and if possible specify detailed format of the expected output data.