The expansion of platform-based informal labor markets across the developing and developed world has produced a class of workers whose contractual precarity is structurally embedded rather than incidental. This paper proposes a graph-theoretic framework to model these labor markets as weighted directed graphs in which nodes represent workers, employers, and intermediary platforms, and edges encode asymmetric relationships of economic dependency, information inequality, and contractual power. We introduce the concept of vulnerability propagation — a process by which adverse shocks such as wage theft, sudden platform deactivation, or unsafe working conditions spread through the network via contagion dynamics analogous to epidemic spreading models. Drawing on structural hole theory, betweenness centrality, and percolation thresholds, we argue that exploitation is not a random occurrence but a topologically predictable event concentrated at specific network positions. The framework offers a principled basis for policy interventions that target the network architecture itself rather than individual workers or firms, and we illustrate its applicability through a case study of platform-mediated delivery labor markets in urban India.