Diversity of stripped-envelope supernova light curves from interactions with binary-driven circumstellar material A growing number of core-collapse supernovae (CCSNe) exhibit diverse light-curve morphologies that indicate strong interactions with dense, pre-existing circumstellar material (CSM). Understanding the physical origin of such CSM is essential for exploring the late-stage evolution of SN progenitors and the diversity of observed transients. Non-conservative mass transfer during binary interactions provides a promising channel for producing dense CSM before core collapse. Using the stellar evolution code MESA, we simulate post-common-envelope binaries that give rise to ultra-stripped progenitors before core collapse, and self-consistently construct the CSM from the mass loss during binary evolution. We further analytically predict the resulting bolometric and radio light curves by following the shock dynamics of SN ejecta interacting with the CSM. We find that surface-radius variations in the ultra-stripped progenitors, which reflect changes in core energy generation after a delay comparable to the thermal timescale, trigger multiple episodes of mass transfer and give rise to a variety of CSM density profiles, including detached shells and multi-peaked structures. When interacting with SN ejecta, such CSMs give rise to non-monotonic, multi-peaked optical and radio light curves that qualitatively resemble features observed in some stripped-envelope SNe. Future long-term multi-wavelength monitoring of ultra-stripped SN candidates, particularly at late times, will be crucial for probing structured CSM and constraining the long-term mass-loss history of stripped progenitors prior to core collapse. Our results suggest that SN-CSM interaction in binaries hosting ultra-stripped progenitors can provide a possible channel for producing diverse light-curve morphologies, highlighting a potential connection between the pre-SN evolution of massive binary stars and the diversity of their explosive transients. 715 EDP Sciences application/pdf