Entanglement in Quantum Vaults and Classical Forges

We treat the vault--forge distinction of the entanglement-tapestry framework as a problem in open quantum systems. Starting from a pre-geometric graph Hamiltonian with a valence potential, a local swapping term, and a combinatorial frustration term, we partition the threads into a dense subsystem (the vault) and a sparse environment (the forge). Tracing out the forge under a standard Born--Markov and secular approximation yields a Lindblad master equation. The new content is not the form of the generator; it is the identification of the jump channels with kinetic swaps that straddle the cut. The leading decoherence rates therefore track the number of cut channels, \(Γγ_0 N∂\), rather than an externally chosen spectral density. High interior density lengthens coherence only when it lowers \(N∂/|C|\) at fixed boundary sparsity. Vaults and forges are the two regimes of one reduced dynamics, not two ontologies. We state the approximations that produce the scaling, identify where they fail, and list measurements in existing platforms that could falsify it. No gravitational dynamics are derived here.