Nutrition‑sensitive agriculture (NSA) is recognised as a pathway to reduce child stunting, with agriculture interventions influencing food access, care practices, and health environments. The FAO framework highlights how on‑farm availability, market food environments, income, women’s empowerment, and nutrition knowledge shape diets and ultimately nutritional status (1). However, most assessments emphasise dietary diversity as a proxy outcome, underscoring the need for a composite index that integrates these domains to directly evaluate impacts on stunting. This study explores how open national datasets can be operationalised into NSA related indicators and examines their association with stunting prevalence. A prototype composite index was constructed using indicators from multiple open sources, drawing on multi‑year data (2020–2024) from 38 provinces in Indonesia. Crop production and livestock population data from Badan Pusat Statistik (BPS) were used to calculate crop diversity (CDI) and livestock diversity (LDI) according to the Shannon Index. Additional BPS data included the Food Insecurity Experience Scale (FIES), women’s income contribution (%), and the percentage of households using safely managed drinking water services. Diet quality was assessed using Minimum Dietary Diversity for Children (MDD‑C), while average daily calorie and protein intake per capita, together with stunting prevalence as the nutritional outcome, were obtained from the Ministry of Health (Kemenkes). Food price stability was represented by the consumer‑level food price variation coefficient from Badan Pangan Nasional. Indicator polarity was standardised (“higher = better”), and weights were optimised to maximise correlation with stunting prevalence. Correlation and linear regression analyses were conducted in R (version 4.3.2). The composite demonstrated a strong negative correlation with stunting (r = –0.73). Weighting analysis identified food access (0.49) and price stability (0.23) as dominant contributors, with crop diversity (0.20), livestock diversity (0.14), and diet quality (0.13) providing additional explanatory power. Regression analysis confirmed the composite as a significant predictor of stunting (β = –53.8, p < 0.001), explaining 53% of variance (R² = 0.53). While most NSA interventions have focused on dietary diversity as the primary outcome (2), this composite demonstrates how upstream pathways such as on‑farm availability, market food environments, and women’s income shape food access and diet, which in turn lead to measurable improvements in nutritional status. This exploratory analysis demonstrates the feasibility of operationalising open Indonesian datasets into a composite framework that links agriculture pathways with child nutrition outcomes. The prototype is validated across provinces and years, but further testing with additional FAO indicators and alternative methods is needed to establish broader reliability and strengthen agriculture–nutrition monitoring frameworks for policy use in Indonesia.