The Techniques and Mechanisms Described Herein are Directed to an Extensible security architecture that provides a security mechanism for minimizing security problems within interpretive environments. The extensible security architecture comprises a script engine configured to process a script and a security manager configured to monitor the processing of the script based on a security policy. The security manager determines whether to open an assembly associated with a command within the script, whether to process the command, whether to allow certain input to the command, and the like. The security policy may be implemented by overriding one or more methods of a base security class that are called when processing the script. The input may be an object passed via an object-based pipeline.