This paper presents the Life Stability Principle (LSP), a framework that conceptualizes emotion, value judgments, and behavior as corrective processes through which living systems maintain stability. Living systems are treated as open systems that maintain dynamic stability through the circulation of matter and energy and the circulation of information; in humans, stability conditions are organized into two functional dimensions, the Physiological Stability Variable (P) and the Relational Stability Variable (R). Emotion is conceptualized as a bodily signal associated with changes in the stability conditions represented by P and R, potentially preceding conscious evaluation, while judgment and behavior are understood as corrective processes that respond to those changes. This framework offers a common basis for comparing the Free-Energy Principle, the somatic marker hypothesis, and prospect theory as complementary existing theoretical accounts, and it indicates possible applications to understanding human behavior and to human-centered AI design. The aim of this paper is to present the conceptual and biophysical skeleton of LSP; mathematical formalization, quantitative predictions, and specific information-processing implementations are addressed in separate papers.