Core-collapse supernovae play a key role in many of astrophysical processes, but the details of how these explosive events work remain elusive. Many questions about the CCSN explosion mechanism and progenitor stars could be answered by either detecting very-low-energy supernovae (VLE SNe) or alternately placing a tight upper bound on their fraction of the CCSN population. However, VLE SNe are by definition dim events. Many VLE SNe result from the failure of the standard CCSN explosion mechanism, meaning that any observable signature must be created by secondary processes either before or during the collapse. In this dissertation I examine alternate means of producing transients in otherwise-failed CCSNe and consider the use of shock breakout flashes to both detect VLE SNe and retrieve progenitor star information. I begin by simulating neutrino-mediated mass loss in CCSNe progenitors to show that a dim, unusual, but still observable transient can be produced. I then simulate shock breakout flashes in VLE SNe for both the purposes of detection as well as extracting information about the exploding star. I discuss particular challenges of modeling shock breakout at low energies and behaviors unique to this regime, in particular the behavior of the spectral temperature. All simulations in this dissertation were done with the CASTRO radiation-hydrodynamic code.