feat(M14): combine continuum AdS double-copy lineage

[jagg-ix]

Combined M14.1–M14.4

This PR supersedes the former stacked PRs #172, #173, and #174 and contains their complete final tree plus M14.4 registration in one commit directly on main.

Scope

• M14.1 causal retarded/advanced Green recurrences, Pauli–Jordan propagation, energy bounds, and positive-frequency Hadamard probe transport;
• M14.2 pointwise Jacobi numerator families, square-summable infinite BCJ direct limit, analytic tail control, color replacement, and generalized-gauge invariance;
• M14.3 D3/boundary-central-charge normalization of the continuum BCJ limit, GKP kernel and Hessian, conformal scaling, RT entropy, and Complex-Einstein normalization;
• M14.4 nested harmonic slabs, compatible Lorentzian metric direct limits, harmonic constraint propagation, fixed-Cauchy uniqueness, expanding-domain Green compatibility, model registration, and documentation.

Formal authority

• repository: jagg-ix/entropic-physlib-private
• branch: entropic-physlib-linear-full
• TIP: ea6c394fcb1d55546d11cd6af3df6556c610d52e

Status boundary

The final result remains a conditional-model. It does not claim an unconditional global or loop-level AdS double-copy theorem, automatic Green/Hadamard existence, an interacting holographically renormalized Witten functional, a nonperturbative quantum-gravity measure, or equality between BCJ amplitudes and RT entropy.

Validation

• all M14 Python modules compiled;
• all four formal JSON ledgers parsed;
• focused tests: 4 passed;
• M14.1 acceptance: 8/8;
• M14.2 acceptance: 9/9;
• M14.3 acceptance: 9/9;
• M14.4 acceptance: 10/10;
• no workflow file was added or modified.

Branch audit

• base: main@6fa3855d60e21c8c48c6fff4cb8eea6948103f14
• head: 28ff3c146f08c5e3c5363011584ed46bfcb097a0
• one commit
• complete M14.1–M14.4 tree
• zero commits behind main

[jagg-ix]

Final audit after publication:

• branch is exactly one commit ahead and zero behind its M14.3 parent;
• final diff is eight files: registry docs, package export, model registration, M14.4 module, theorem ledger, ZIL graph, and focused test;
• local exact-tree validation: Python compilation passed, all four M14 ledgers parsed, and pytest -q tests/test_m14_*.py passed 4/4;
• acceptance groups: M14.1 8/8, M14.2 9/9, M14.3 9/9, M14.4 10/10;
• GitHub reports the PR mergeable;
• existing workflow run 30587894472 succeeded. It is the repository's M11 path-filtered compile/ledger/test job, not a dedicated M14 test job. The M14-specific evidence is the focused local four-test campaign recorded above.

The theorem status remains conditional-model; no unconditional global or loop-level continuum AdS double-copy theorem is claimed.