Proto-Pineal v7.5: Demonstrating Trajectory Sensitivity Beyond Endpoint State in a Deterministic System

Proto-Pineal v7. 5 is a deterministic discrete-time substrate for studying lawful state evolution under fixed regulatory constraints. The system is not an agent architecture and does not assign decision authority to external components such as language models. All state transitions occur through a single canonical update path enforced by ordered regulators and bounded admissibility conditions. This release introduces a trajectory-oriented observation layer over replayed runs. A canonical post-run object (TrajectoryRecord v1) captures ordered state evolution, including identity drift, energy, spectral measures, basin regimes, and derived trajectory observables. A replay-safe counterfactual assay is provided to test trajectory sensitivity. Two runs are constructed from a shared genesis and converge to near-identical endpoint states while differing in ordered input structure. Despite endpoint convergence, the runs exhibit strong divergence in trajectory geometry, including second-order curvature (Kₜ) and derived trajectory metrics. This demonstrates that, within the system, ordered path information is not reducible to endpoint state. The result establishes a minimal, falsifiable surface for studying trajectory-sensitive dynamics in deterministic systems. The release includes: the full v7. 5 paper (PDF) trajectory sensitivity assay artifacts (plots and JSON outputs) sample trajectory records a manifest with integrity hashes a reference to the corresponding repository commit No claims of cognition, agency, or semantic reasoning are made. The contribution is restricted to the definition and demonstration of trajectory-sensitive structure in a regulated dynamical substrate under deterministic replay. This version is intended as a stable, auditable baseline for further investigation of history-dependent behavior and trajectory-level observables in synthetic systems.