Corporate financial management requires a recurring allocation of scarce internal funds among liquidity reserves, working capital, debt reduction, and strategic investment. These uses are usually evaluated in separate analytical routines, even though their value is jointly determined by financing frictions, cash-flow uncertainty, and downside exposure. This study develops an integrated Financial Resilience Index (FRI) and a scenario-based allocation model that combines expected financial benefit with Conditional Value-at-Risk (CVaR) of target shortfall. The framework also accommodates Explainable Artificial Intelligence (XAI) when machine-learning forecasts are used to estimate scenario payoffs: driver-level explanations can be generated with SHapley Additive exPlanations (SHAP) before the optimization stage, while the allocation model itself remains transparent. Because no observational dataset is assumed, the paper presents a fully disclosed synthetic decision experiment rather than empirical claims. In the illustration, a growth-first policy yields the highest expected benefit but produces severe downside losses, whereas a balanced policy reduces CVaR materially and a resilience-first policy lowers downside shortfall CVaR by approximately 97% relative to the growth-first solution while sacrificing about 20% of expected benefit. The framework therefore identifies an interpretable resilience frontier instead of a single mechanically “optimal” cash policy. The principal contribution is to connect working-capital efficiency, corporate liquidity, debt capacity, and strategic investment within one auditable financial-management architecture, allowing chief financial officers to choose an allocation consistent with explicit risk appetite and liquidity floors. The framework is intended as a decision-support model that can subsequently be calibrated and tested with firm-level accounting, treasury, and cash-flow data.