Generative AI and design as a bridge for rural craft revitalization: a workflow-centric framework

Rural handicrafts in China face a persistent “value gap” between local craft practices and contemporary aesthetic, emotional, and functional expectations. While generative AI has increasingly been introduced into craft revitalization, less is known about how algorithmic outputs are translated into materially feasible making processes, and how such workflows can remain traceable, inspectable, and locally accountable. This study examines a 2025 field experiment in Shengshiying Village, Yongning Town, Yanqing District, Beijing, a district-level intangible cultural heritage site for feather painting. Drawing on interviews, participatory workshops, prompt/version logs, material records, and artefact documentation, the paper proposes and empirically examines a workflow-centric, provenance-aware AIGC+design framework for rural craft revitalization. The framework traces how craft knowledge capture, motif generation, template preparation, material matching, guided assembly, and local coordination were connected through documented handoffs. The intervention produced a 320-entry material database, enabled bounded-session motif exploration, generated 600+ motif variants, supported screening for assembly feasibility, and trained 15 local makers and 5 guides. Process records also indicate reductions in motif preparation time and novice material-matching effort under standardized templates and logged selection criteria. Rather than presenting a universal model of rural revitalization, the study offers a pilot-derived account of how production-side constraints and perceived-value formation can be linked through an accountable craft workflow. It further summarizes this account through a Cultural Activation Path Matrix observed in the Shengshiying pilot and specifies boundary conditions for transfer, including local idiom anchoring, a coordination node, and documentation discipline. Limitations include the single-site, time-limited pilot and the need for longitudinal evidence on authenticity, benefit sharing, and stewardship of a shared digital commons.