This paper presents several enhancements to ARIFS, a reuse environment that sets the foundations for reusing formal verification efforts in an iterative and incremental software process for the design of distributed reactive systems. A criterion based on generic components is added, together with a self-learning mechanism, to reduce the search space and maximize the probability of retrieving useful information. Besides, a formalization is given on how to apply verification tasks on a reduced number of states when the retrieved information is not enough for the user's intents. These enhancements are shown to improve both the effectiveness and the efficiency of ARIFS.