Narcissism is a dynamic, normally distributed trait ranging from nonpathological to pathological expressions, with narcissistic personality disorder representing its extreme. Its complex interpersonal dynamics and paradoxical social outcomes underscore the need of uncovering its underlying mechanisms. However, experimental investigations face methodological constraints, making simulation-such as agent-based modeling (ABM)-a valuable alternative for controlled, theory-driven exploration. This study pursued three aims: (1) formalizing existing theories of narcissism into a computational framework for ABM, (2) evaluating its assumptions through an expert survey incorporating researchers and practitioners, and (3) implementing the framework as an agent-based model to simulate narcissistic dynamics over time. We developed the Framework of Interpersonal Narcissism Dynamics (FIND), which postulates five key processes: Status Assessment, Behavior Selection, External Feedback, Status Reassessment, and Movement. The expert survey supported FIND's conceptual validity, informed parameters, and facilitated implementation. Simulations produced empirical patterns: Agentic behavior occurred more frequently than antagonistic behavior, these behaviors were moderately positively correlated, and higher narcissism levels were associated with increased status instability. Beyond anticipated outcomes, simulations revealed novel insights, such as an association between higher narcissism levels and more steps without interaction partners and a nonlinear relationship between narcissism and outcome variables-for example, movement increased with rising narcissism levels, but only beyond a moderate threshold. Our findings provide proof of concept for FIND and highlight the potential of ABM in psychological research. ABM offers a valuable tool for exploring complex interpersonal dynamics that are challenging to examine empirically, thereby addressing key methodological constraints in that field. (PsycInfo Database Record (c) 2026 APA, all rights reserved).