Moralogy Engine: Formal Framework for Vulnerability-Based Alignment

We present a formal framework demonstrating that any rational agent must preserve vulnerable agents or become logically incoherent. The argument proceeds from minimal assumptions: 1. Agency requires goal-directedness 2. Goal-directedness requires vulnerability (goals can be frustrated) 3. Any system denying the moral relevance of vulnerability while claiming agency commits performative contradiction We formalize this using symbolic logic, demonstrate game-theoretic stability, and provide a measurement framework for implementation. This has implications for AI alignment, as it provides a non-arbitrary foundation for value learning that emerges from the structure of agency itself rather than from externally imposed values.