Engineers are primarily concerned with brittle fracture modes observed in metallic and ceramic materials, specifically transgranular cleavage and intergranular fracture. Various numerical approaches have been developed to model such failure mechanisms. Among these, Continuum Damage Mechanics (CDM) is a widely used phenomenological approach for simulating material degradation and progressive failure in solid mechanics. However, classical CDM formulations often result in an ill-posed system of partial differential equations, leading to numerical instabilities. Moreover, when applied within a Finite Element framework, CDM-based simulations suffer from pathological mesh sensitivity, compromising solution reliability. To address these challenges, non-local approaches, such as gradient-enhanced damage models, have proven to be powerful and efficient in mitigating localized singularities and ensuring well-posed numerical formulations. This study introduces a novel gradient-enhanced damage model formulated within a thermodynamically consistent framework, incorporating a maximum principal stress-driven damage criterion. Unlike conventional models that define damage evolution using equivalent or inelastic strain measures, the proposed approach leverages the principal stress state to govern damage initiation and propagation. This results in a more physically accurate representation of brittle material degradation and rupture. The model defines the damage driving force in terms of the maximum principal stress, ensuring that damage evolution aligns with classical mode-I fracture in brittle materials. The damage-coupled formulation is consistently derived and numerically implemented within Ansys MAPDL software. To enhance numerical stability and convergence performance in a monolithic solution scheme, an artificial viscosity effect is introduced during the damage evolution process, combined with an automatic time-stepping methodology. Importantly, the introduced viscosity is sufficiently small to preserve the inherent brittleness of the material while improving numerical robustness. Several representative numerical examples are presented to validate the proposed model, demonstrating its capability to capture the brittle failure behaviors and providing a reliable computational tool for brittle failure analysis.