CAPEC-160 Exploit Script-Based APIs

Some APIs support scripting instructions as arguments. Methods that take scripted instructions (or references to scripted instructions) can be very flexible and powerful. However, if an attacker can specify the script that serves as input to these methods they can gain access to a great deal of functionality. For example, HTML pages support <script> tags that allow scripting languages to be embedded in the page and then interpreted by the receiving web browser. If the content provider is malicious, these scripts can compromise the client application. Some applications may even execute the scripts under their own identity (rather than the identity of the user providing the script) which can allow attackers to perform activities that would otherwise be denied to them.

1. [Explore] [Identify API] Discover an API of interest by exploring application documentation or observing responses to API calls 2. [Experiment] [Test simple script] Adversaries will attempt to give a smaller script as input to the API, such as simply printing to the console, to see if the attack is viable. 3. [Exploit] [Give malicious scripting instructions to API] Adversaries will now craft custom scripts to do malicious behavior. Depending on the setup of the application this script could be run with user or admin level priveleges.

• The target application must include the use of APIs that execute scripts. • The target application must allow the attacker to provide some or all of the arguments to one of these script interpretation methods and must fail to adequately filter these arguments for dangerous or unwanted script commands.