# n66 — pre-registration: the self-referential ledger (before any code)

**The life-or-death gate of Drainage II.** Headstone 23 measured
beta = 0.513: the drainage field's own maintenance does not gravitate.
The completion's first tissue: THE COST OF AN ENTRY IS ITSELF AN ENTRY —
field cost density added to consumption, at conversion kappa LOCKED TO 1
(one ledger currency; the attack record disarmed the dial in advance —
any kappa != 1 outcome is recorded as a fit, i.e. a failure).

## Construction (declared)
- Local cost density: rho(i) = (1/2) * (1/2) * sum_{edges e incident i}
  dk_e (G dk)_e — the per-site share of the total cost (sums to it
  exactly; G's measured near-locality makes the split principled).
- Self-consistent drainage: iterate dk_{n+1} = drain(G', c_matter +
  rho[dk_n]) to a fixed point; G' = info metric at the deformed vacuum
  as in n65 (the state dependence stays on).
- beta extraction: exactly the n65 verified procedure (E1, E2 by
  F-antisymmetrization; beta = 1/2 + E2 E_t0 / E1^2; F-scan for the
  weak-field limit) with the self-referential fields throughout.

## Gates
- N1 (attack-4 gate, convergence): the iteration converges (relative
  update < 1e-10) within 50 iterations at every F used; report the
  contraction factor. Divergence at weak field = the completion dies
  by runaway.
- N2 (THE NUMBER): beta_toy, F-scan as in the verification pass
  (F = 0.3 ... 0.0375). LIFE: beta -> 1.000 within 2% in the weak-field
  limit, with kappa = 1 untouched. DEATH: any other stable value —
  the completion's central mechanism fails and Drainage II is buried
  with headstone 23's own knife, before its dynamics was ever built.
  PARTIAL, declared possible: beta moves from 0.513 toward 1 but lands
  elsewhere (e.g. the 1D share convention biases the coefficient) —
  recorded as shape-pass/value-fail, one registered convention check
  allowed (the rho split), nothing else.
- N3 (gamma untouched): the first-order sector must not move —
  E1's linear coefficient within 0.5% of its n65 value (the fix must
  not break what worked).

## Honesty clauses
- 1D, static, one rho-split convention (declared above); the
  self-sourcing of KINETIC/current sectors is out of scope (that is
  the dynamics tissue).
- The iteration adds net positive consumption; the compact-ring
  jellium (lstsq) treatment absorbs it as everywhere else.

## RESULTS (same day; n66_self_ledger.py) — HEADSTONE 24
- N1 PASS: convergence in 3-4 iterations at every F (no runaway;
  attack 4 disarmed).
- N3 PASS: E1 shifted 0.24% — the first-order sector untouched.
- N2: DEATH. beta = 0.5130 at every F — the self-sourcing changed
  NOTHING (0.513 -> 0.513 to the fourth digit).
- DIAGNOSIS, verified numerically: the field cost density is
  PARITY-EVEN BY CONSTRUCTION (a cost has no sign to alternate); its
  staggered charge is exactly 0.0e0, and the field it sources is
  1000x weaker than matter's and short-ranged. In the bipartite
  architecture, gravitational charge lives in the staggered sector —
  SO FIELD ENERGY IS GRAVITATIONALLY NEUTRAL, structurally. beta is
  locked at 1/2 + (state-dependence, 0.013) no matter what is fed
  back. Combined with Theorem 2 (signed architecture: smooth charges
  CAN source but like charges repel), the dichotomy acquires a second
  face, THEOREM-4 GRADE (proof sketch = this measurement + the n51a
  filter argument): IN QUADRATIC DRAINAGE, ATTRACTION AND beta = 1
  EXCLUDE EACH OTHER — the same pole structure that blocked the
  clocks blocks the field's energy from gravitating.
- SCOPE, stated precisely: n66 kills "beta from the audited scalar
  sector" — the completion's tissue 2 AS DESIGNED. In general
  relativity the nonlinearity lives in the tensor sector's
  self-coupling; Drainage II's remaining route to Mercury is therefore
  entirely through tissue 3 — the free TT sector and year 1's formal
  chain (TT coupling -> unimodular gravity -> Bianchi + conservation ->
  FULL Einstein, whose beta = 1 is a theorem of that chain, not a toy
  measurement). The completion's beta burden shifts from computation
  to the formal TT junction: either the junction holds and beta = 1
  arrives with the tensor sector, or Drainage II has no Mercury and
  the funeral is total. No third route remains.