Digital watermarking has become very important for copyright protection and authentication. Broad claims have been made about the `robustness' of various digital watermarking method. Unfortunately the criteria as well as the pictures used to demonstrate these claims vary from one system to the other and recent attacks show that the robustness criteria used until now are inadequate. This paper shows that majority of attacks implemented in the Stirmark package can be done in a quick and easy way using Octave. Given a watermarked image, StirMark will apply several transformations with various parameters then produce an output image that can be tested using benchmark to judge the robustness of image watermarking systems.