This systematic characterisation examines how urban-level resource-use models represent agents and their capacity to influence metabolic change. It distinguishes between models that ignore agency, treat actors as external parameters or engage more directly with decisions, behaviour and institutional action.

The study reviews model structures and classifies their levels of engagement with agents, clarifying what kinds of questions each approach can answer. The comparison reveals where technically detailed resource assessments remain weak in explaining transition processes and where stronger links to social and decision sciences are needed.

The article advances the aIST-nexus agenda by turning agency into an explicit criterion for model evaluation. It helps researchers select or develop models according to whether the objective is accounting, scenario exploration, explanation or decision support.