The large number of historic masonry churches exposed to seismic hazard, together with the need for rapid, consistent and scalable assessment procedures, highlights the limits of current practices when applied at a territorial scale. Seismic vulnerability assessment is often constrained by time, data fragmentation and the difficulty of ensuring uniform evaluations across different survey campaigns and surveyors. A digital assistant, called Masonry building Assessment of Seismic vulnerability based on Observed damage and Neural networks (MASON), is developed to support the assessment of seismic damage and vulnerability of historic masonry churches through a structured and operational workflow. The tool, thoroughly described in the present paper, supports the entire process, enabling the collection of survey data and the interactive construction of simplified 3d models on-site. The framework is grounded in the Italian assessment practice for masonry churches and is developed around the second-level A-DC form and the Italian Guidelines for the evaluation and reduction of seismic risk of cultural heritage. It adopts the macro-element approach used for churches [1] and relies on a large database of observed seismic damage from past earthquakes, including the L’Aquila earthquake (2009), the Central Italy earthquakes (2016–2017) and the Ischia earthquake (2017). Survey data, typological classification and geometric information are combined to support the identification of recurrent failure mechanisms and to associate the surveyed church with specific vulnerability models. Data-driven components aid typology recognition and vulnerability estimation, complementing expert-based assessment criteria. The proposed tool is designed as a decision-support system aimed at improving consistency and efficiency of seismic damage and vulnerability assessment, while remaining compatible with established methodologies. Although developed within the Italian context, the approach is intended to be adaptable to different territorial settings and seismic scenarios, supporting large-scale applications.