A non-aromatic prodrug candidate for endometriosis, designated END-102, is reported. The design eliminates aromatic rings entirely, replacing the conventional heterocyclic scaffold with a trans-4-aminocyclohexane core bearing a urea linker, an aminodiol pharmacophore arm, and an oxetane hydrogen-bond acceptor. The prodrug form (END-102b) is activated in vivo via sequential cytochrome P450 3A4-mediated O-demethylation and N-demethylation to the active species (END-102a). Computational binding calculations predict a dissociation constant of 17 nM at the progesterone receptor B (PR-B) ligand-binding domain, with 400-fold selectivity over aromatase (CYP19A1). Multi-framework safety profiling was conducted using the Cascade Interaction Liability Model (CELM), the Human Analogue 100-slot virtual Phase I simulator, and the Human Gut Model v1.0 72-slot gastrointestinal compartment simulator. The active form achieves a CELM score of 0.00 (class: SAFE). Virtual Phase I assigns a verdict of PROMISING with a Boltzmann selectivity of 83,226-fold and zero chain reactions across 100 biological slots. Gut model analysis identifies PEPT1/PEPT2-mediated uptake as the primary absorption mechanism. No mucosal toxicity, microbiome, or reactive metabolite flags are detected. END-102 represents a structurally clean, pharmacologically selective lead candidate appropriate for advancement to in vitro validation.