Direct Static Code Injection
Last revision: 04/7/2009
A Direct Static Code Injection attack consists of injecting code directly onto the resource used by application while processing a user request. This is normally performed by tampering libraries and template files which are created based on user input without proper data sanitization. Upon a user request to the modified resource, the actions defined in it will be executed at server side in the context of web server process.
Server Side Includes is considered a type of direct static code injection. It should not be confused with other types of code injection, like XSS (“Cross-site scripting” or “HTML injection”) where the code is executed on the client side.
This is a simple example of exploitation of a CGISCRIPT.NET csSearch 2.3 vulnerability, published on Bugtraq ID: 4368.
By requesting the following URL to the server, it’s possible to execute commands defined on the ‘’’’setup’’’ variable.
For the classic example, the following command can be used to remove all files from “/” folder: csSearch.cgi?command=savesetup&setup=`rm%20-rf%20/`
Note that the above command must be encoded in order to be accepted.
This example exploits a vulnerability on Ultimate PHP Board (UPB) 1.9 (CVE-2003-0395), which allows an attacker to execute random php code. This happens because some user variables, like IP address and User-Agent, are stored in a file that is used by the admin_iplog.php page to show user statistics. When an administrator browses this page, the previously injected code by a malicious request is executed. The following example stores a malicious PHP code that will deface the index.html page when an administrator browses admin_iplog.php.
GET /board/index.php HTTP/1.0 User-Agent: <? system( "echo \'hacked\' > ../index.html" ); ?>