This paper presents a software tool to deal with the diagnosis of discrete-event system modeled by labeled Petri nets (LPNs). This tool, called PEtri-Net-based Diagnosability Analyzer (PENDA), is an updated version of the prototype that implements our techniques for (K-)diagnosability analysis of discrete-event models. In order to cope with the state explosion problem that arises when establishing a diagnoser, PENDA implements an incremental and on-the-fly algorithm to analyze (K-)diagnosability while partially building the state space of the LPN. In addition, several improvements have been carried out in the source code to improve the time/memory efficiency of the analysis. The current version shows better performance and allows for handling LPN models encoded in the standard PNML format.