Pattern Agency on Active Media (PAAM) 8.0.3 develops a mathematical theory of localized patterns, their stability, interaction, persistence, and agency in noisy dynamical media. The theory is constructed from a single foundational axiom: a stochastic substrate represented by a diffusion on a finite-dimensional smooth manifold with a nondegenerate noise tensor. A medium is obtained by imposing locality, homogeneity, and a symmetry group compatible with the drift and noise; its homogeneous equilibria define the background. Patterns arise as localized symmetry-breaking states whose deviation from the background decays away from a centre with a scale-independent spatial rate. PAAM derives the structure and dynamics of such patterns from the geometry and spectrum of the underlying stochastic system rather than introducing an independent notion of individuality. The linearisation separates neutral modes associated with symmetry, internal pattern modes, and extended modes of the background. A spectral gap between internal and background modes provides a criterion for treating a localized structure as an individuated pattern. The theory further formalizes perturbation response through response functions, pattern pose and its stochastic diffusion, return-map invariants, stability boundaries, and fluctuation growth near fold, flip, and Neimark–Sacker instabilities. PAAM distinguishes two notions of energetic or informational cost: path irreversibility supplied by the stochastic substrate and thermodynamic dissipation requiring an explicitly declared physical embedding. On this basis, persistence, maintenance, and agency are defined without identifying agency with irreversibility alone. The framework also treats pattern birth, escape, dissolution, fading, division, lineage, interaction, binding, pinning, and motion, including their dependence on noise, control parameters, symmetry, and background dynamics. The theory is formulated to separate mathematical consequences of the substrate and medium from hypotheses, model-dependent embeddings, and experimentally testable claims. It therefore provides a common mathematical language for analysing localized structures across stochastic physical, chemical, biological, and other dynamical systems, while making explicit which properties follow from the formalism and which require additional empirical assumptions.