Developed for the Principles of Complex Systems (PoCS) class as part of the Complex Systems and Data Science Certificate at the University of Vermont (UVM), this project introduces a novel computational framework for narrative analysis. Graphic Stories leverages Large Language Models (LLMs) to automatically decompose narrative texts into two dynamic streams: character interaction networks and "archetypometric" personality trajectories. By visualizing these structures in an interactive dashboard, the system reveals how character psychological evolution drives plot mechanics, grounding abstract narrative theory in quantitative, observable data.