Rural low-carbon planning requires evaluation tools that connect carbon-related performance with the ecological, socioeconomic, and infrastructure conditions of rural settlements, yet carbon inventories and conventional human-settlement assessments do not fully integrate these dimensions. This study developed an exploratory carbon-benefit-oriented sustainability diagnostic and applied it to Qingge Village, Fujian Province. Fuzzy Delphi screening retained 27 indicators across ecological, social, production, and engineering subsystems; Delphi consultation generated expert-based weights; a dimensionless ratio transformation standardized the indicator values; threshold, standardization, and weighting sensitivity analyses evaluated the stability of results across analytical specifications. The primary composite score was 0.413 compared with 0.590 under the target-achievement transformation, and both expert-weighted transformations ranked subsystem contributions as production, ecological, social, and engineering. The ecological and production subsystems performed more strongly against the selected references than the social and engineering subsystems, whereas fully equal global weights placed ecology ahead of production. The principal diagnostic priorities were household energy emissions, green agricultural upgrading, public-transport access, and recorded energy-efficient building coverage. These findings indicate that the diagnostic can organize village-scale low-carbon planning priorities and translate multidimensional performance evidence into a sequenced set of spatial improvement options. This study therefore provides a transparent and practically applicable approach for linking carbon-related indicators with rural human-settlement planning and supporting evidence-informed low-carbon revitalization.

Exposure to high-quality visual environments characterized by features such as structural order, biophilic/natural elements, and positive atmosphere is important for walking experience within community life circles (CLCs). However, compared with the static configuration of visual environments within CLCs, dynamic walking-based exposure may highlight unpredictable encounter areas and heterogeneous environmental quality. Neglecting this mismatch may misdirect environmental interventions and limit their health-promoting potential. This study aims to integrate multidimensional visual environment features into interpretable scene clusters to improve comparability and measure configuration–exposure mismatches across CLCs at scale. We examine 1262 CLCs in Harbin, China, identifying visual scene clusters from 67,840 street-view images and extracting exposure frequencies from 981,500 mobility tracks. The results show that (1) eight scene clusters effectively describe the complex visual environments of CLCs; (2) significant small-to-moderate mismatches exist between configuration and exposure; (3) the trend of commute-related walking activity is often consistent with strengthened exposure to high-disorder scenes and weakened exposure to some high-quality scenes with positive atmospheres. This study provides a data-driven framework for identifying mismatches in both the intensity and spatial distribution of visual scene configuration and exposure, supporting refined community environmental governance.