As cislunar activity accelerates (Artemis, commercial relays, navigation beacons), the Earth–Moon system risks repeating the congestion and “design-now, mitigate-later” pattern seen in LEO. This project developed a quantitative, early-design framework that treats end-of-life (EOL) disposal feasibility as a first-class objective in cislunar spacecraft constellation design, rather than an operational afterthought. The core contribution is a novel End-of-Life Capability Metric (ECM), a dimensionless sustainability indicator that scores a constellation architecture by combining: (1) the fraction of satellites that can successfully reach a safe terminal state (2) the propellant efficiency of disposal maneuvers (3) the aggregated risk of execution failure, post-disposal containment loss, and collision exposure. This ECM is embedded into a multi-objective optimization formulation alongside traditional objectives such as coverage, orbit stability, and spacecraft count/cost, enabling explicit trade-offs between mission performance and long-term stewardship. Methodologically, the framework unifies three modeling domains: CR3BP dynamical propagation (to capture Earth–Moon multi-body effects and instability pathways), constellation architecture evaluation (coverage and stability), and EOL strategy modeling across disposal pathways such as lunar impact, heliocentric escape, and cislunar graveyard options. The implementation leverages high-fidelity periodic orbit initial conditions (via the JPL three-body periodic orbit database) and uses systematic trade-space exploration to identify architectures that maintain performance while improving disposal robustness. Case studies show that including EOL capability can materially change the “best” constellation designs, often achieving comparable coverage with reduced long-term risk and minimal additional cost.