The rapid growth of hardware input devices over the last few decades has brought about the challenge of creating compelling interface techniques making use of these technologies. Research has thus been focused on reducing a given set of input events into a particular interactive application. However, taking input event parsing and implementing user-defined actions as separated and well defined independent tasks can provide several advantages, such as easing maintenance and extensibility of interactive environments, while providing a common background for user interaction comparative studies. Our key idea for implementing user-defined actions on application objects is to be able to bind such actions with the degrees of freedom (DOF) needed to carry them out. We introduce the design and implementation of Proscene-3, an open source framework comprising three layers: a low-level component providing a set of virtual events (bogus-events) and the means to bind user-defined actions to them, thereby allowing separating the two aforementioned tasks; a mid-level component, implementing a feature-rich set of widely-used motion actions allowing picking & manipulation of objects, including the scene viewpoint; and, a high-level library, exposing those features to the Processing language. The results have shown the feasibility of our approach and the expressiveness of the Proscene-3 framework API.