Static code analyzers are helpful in uncovering the potential errors at early stage of developmental life cycle of the project. Every project defines its own naming convention rules and other custom protocols that need to be adhered to in the source code. To increase the usability of code analysers in this scenario, they should support easy integration of project specific rules. In this talk, two approaches for designing extensible static code analyzers are presented. In first approach, the project specific policies are defined separately, in LISP like expressions, from rule analysis engine. At the time of execution, rule analysis engine pick up the policies according to the user's requirement and execute them. Another approach is, to teach the code analyser about the project specific rules by guided presentation of correct and wrong usage of the language pragmatics. User's input is necessary during the learning phase of the code analyser so as to reduce the learning time.