MCA: dense-core owner substrate and reserve repricing (post-#1168)

[AllenGrahamHart]

Stack and review boundary

Stacked on exact #1168 head 6a5dcdae1591fc7f044eda6a942bfe178521a48c (itself on #1167/#1166/#1165). Review only:

6a5dcdae1591fc7f044eda6a942bfe178521a48c..b4bad860750f91955dbaead8f2b5a0fdef1f1343

What this supplies

Four guarded inputs for the owner route that #1168 leaves open (a chronology-correct owner for dense parallel pair cores). It does not construct that owner; its purpose is to make the next theorem type-correct and honestly priced before any dense pair-core group is charged to the S/A/E chronology.

1. Separate two-anchor reserve repricing. The near-rational set and the order-32 exception set cannot be identified (2w = 134944 KoalaBear / 134896 Mersenne-31 vs the reserved 31). Any owner reusing the large rational-owner assembly must meet B_owner^(2w)(g) <= B*-(2w+31)-(n-g), the four charges closing to B* identically; exact endpoint tables for both rows (targets 274980728110346481..274980728111260112 and 15728609..16642288). Arithmetic viability only — the stricter large-owner maximum-fiber theorem is explicitly not claimed.
2. Silent k -> k+1 transport is false. On the deployed domain, (u,v)=(1_E, X^k) at slope zero is support-wise MCA-bad for RS[F,D,k] and simultaneously explained for RS[F,D,k+1]: badness and first-owner semantics are not dimension-invariant. Shipped as a hostile regression.
3. The exact repair. A degree-guarded shifted-lattice adapter (envelope plus code-shift cap) in exact bijection with degree-<k explanations on the identical support, plus an executable same-support pair-noncontainment test.
4. A typed deployed pole-line witness at d1 = 67473 (both shifts), passing the guard and reconstructing its support and explanation, with Q, BC, and U_new retained as UNASSIGNED.

Acceptance contract (the bridge to #1168 §6)

Any S/A/E chronology owner for the forced delta <= 4, 200632-slope terminal must: meet the repriced target (R1); preserve the actual line/support/slope chronology; pass the degree and pair tests of item 3; accept the typed pole line of item 4; and reject the transport mutation of item 2. This is an interface, not an owner claim.

Verification

• primary verifier: PASS (36,015 exhaustive toy records; official endpoint tables; d1=67473)
• independent verifier: PASS (packet, reserve, transport, adapter, pole, bridge; 4/4 controls)
• hostile mutations: 8/8 rejected
• TeX build: 123 pages, clean
• all public-DAG source pins confirmed reachable before pinning
• coordinator replay at the pushed head under RAMguard: PASS

Scope and nonclaims

Zero deployed ledger movement; rank eleven is not paid; no K3, LIST, or closure claim. #1167's rank-10 payment is noted neutrally: the same theorem was reached independently on the public prize DAG (cycle 232) — convergence, not contest.

🤖 Generated with Claude Code

https://claude.ai/code/session_01Ve3d3scwbLSLeo6ByidPxF