HCDD: Decision Boundaries, Liquidity Headroom and Executable Responses. A reproducible diagnostic for capital commitments under uncertainty

Scenario outputs become decision evidence when they are evaluated against explicit objectives, constraints and available actions. This paper specifies a public deterministic core of the Horizon-Conditioned Decision Diagnostic (HCDD). It separates three objects: a preference boundary between competing policies, a constraint boundary beyond which a policy becomes unacceptable, and an execution boundary beyond which a specified response can no longer preserve feasibility. These objects need not coincide. An entirely synthetic adaptation example provides closed-form boundaries, cash-flow paths and a model-version replay. At a fixed upgrade cost, the illustrative loss-rate thresholds for preference reversal, unprotected liquidity exhaustion and loss of upgrade feasibility at a specified review date are 1.3464, 1.4500 and 1.7000 million currency units per year. A timing case shows why subtracting construction time from an unmitigated breach time can substantially overstate the available action window when the intervention itself consumes liquidity. Elementary propositions establish the effects of shared buffers, loss sequencing and approximation error. The resulting specification connects numerical analysis to a decision record containing assumptions, monitoring triggers, action ownership and revision rules. The contribution is an explicit, reproducible integration for conditional decision diagnostics; the illustration is not an empirical validation or an optimal-stopping solution.