Natural hazards pose significant risks to human lives, infrastructure, and ecosystems. Understanding risks along all these dimensions is critical for effective adaptation planning and risk management. However, climate risk assessments mostly focus on population, economic asset values, and road or building infrastructure, because publicly available data on more diverse exposures are scarce. The increasing availability of crowd-sourced geospatial data, notably from OpenStreetMap, opens up a novel means for assessing climate risk to a large range of physical assets. To this end, we present a stand-alone, lightweight, and highly flexible Python-based OpenStreetMap data extraction tool: OSM-flex. To demonstrate the potential and limitations of OpenStreetMap data for risk assessments, we couple OSM-flex to the open-source natural hazard risk assessment platform CLIMADA and compute winter storm risk and event impacts from winter storm Lothar across Switzerland to forests, UNESCO heritage sites, railways, healthcare facilities, and airports. Contrasting spatial patterns of risks on such less conventional exposure layers with more traditional risk metrics (asset damages and affected population) reveals that risk hot-spots are inhomogeneously and distinctly distributed. For instance, impacts on forestry are mostly expected in Western Switzerland in the Jura mountain chain, whereas economic asset damages are concentrated in the urbanized regions around Basel and Zurich and certain train lines may be most often affected in Central Switzerland and alpine valleys. This study aims to highlight the importance of conducting multi-faceted and high-resolution climate risk assessments and provides researchers, practitioners, and decision-makers with potential open-source software tools and data suggestions for doing so.