Register discipline. Every cosmological word here is
register-2: overlays stated + audited, never asserted. THE FIREWALL binds
every panel: no claim about physical cosmology or astronomy, ever; no observational dataset
as input, ever; NO claim on any Millennium Prize problem, ever (rix §8, transitive)
— every surface carries its unknowing statement. The received record enters ONLY
through its published surface (THE HARVEST, 940891b → the record at
7fbbd6a); pinned rix quantities are validity gates, never discoveries
(§5.1). MTX-A1 CLOSED AS WITNESSED — OPEN-EDGE (seal
4a7dddc): the encodings froze at P0; the instrument GROUNDED (the calibration
re-witness 22/22 EXACT); both LINEAR cards carried (the ordered-enrichment direction
3 ≥ 2 · the direct-line coincidence sixteen-for-sixteen); THE TWO-FORM ATLAS
KILLED at the fine grain — 8222 = ALL-BUT-101 at v₂ = 1
(C3’s kill clause named it: the form family’s next boundary); not-copies
carried (2 ≠ 3). MTX-A2 CLOSED AS WITNESSED — OPEN-EDGE (seal
04320aa, THE SIGHTING ATTEMPT): the seven-bit form is a POPULATION
— {4130, 8222, 8234}, all v₂ = 1 (each isolated at its own
±4); THE ORDERING INVERTED at k = 16 (rich5₁₆ = 7 >
rich6₁₆ = 6 — held at k = 8, grain-bounded, the received
lesson-shape); the direct-line coincidence UNBROKEN at 32/32 (the held
law-candidate; no promotion before a declared gate); the 32-cell census QUIET 19
· {000, 110} 10 · ALL-BUT-101 3. MTX-A3 CLOSED AS WITNESSED —
OPEN-EDGE (seal c47ab3b, THE POPULATION ATTEMPT): the seven-bit
family is v₂ = 1-CONFINED at this grain (zero members at v₂ = 2; zero
new members on the home depth — the population stays THREE); the ordering
TIED at k = 24 (9 = 9; the arc 3 > 2 → 7 > 6 → 9 = 9 — the
single-crossing shape dead here); a RICH 4 = 4 at v₂ = 2 (the eras land
as copies in count at the new depth; v₂ = 2 reads DENSER, 8/16 vs 18/48);
the direct-line coincidence UNBROKEN at 64/64 across two depths — the
one survivor; every structural, form and ordering regularity has proven LOCAL. The
64-cell census QUIET 38 · {000, 110} 23 · ALL-BUT-101 3.
MTX-A4 CLOSED AS WITNESSED — OPEN-EDGE (seal f8fd71b, THE
PROFILE ATTEMPT): THE PROFILE HAS A KNEE — the matched first-sixteen depth
profile reads 5 → 8 → 5, a RIDGE peaked at v₂ = 2 (the deepening
card refused; the 5 = 5 flanks a flag); THE TIE BROKE — rich5₃₂
= 13 ≠ 14 = rich6₃₂, era six leads again: the arc 3 > 2 → 7 >
6 → 9 = 9 → 13 < 14 — THE OSCILLATION CONFIRMED; 8 ≠ 7 at
v₂ = 2 (the equality was the sample); THE POPULATION IS FIVE over TWO
depths — {4130, 8222, 8234, 8294} at v₂ = 1 · {4220} at
v₂ = 2 (the A3 confinement was itself grain-local); the direct-line
coincidence UNBROKEN at 112/112 across three depths. The 112-cell census QUIET 65
· {000, 110} 42 · ALL-BUT-101 5. MTX-A5 CLOSED AS WITNESSED —
RESOLVED, THE CAMPAIGN’S FIRST FULL CONJUNCTION (seal a731aec,
THE BAND ATTEMPT): the ridge is ONE-PEAKED — the matched four-depth
profile 5 → 8 → 5 → 5 (flat flanks around the peak);
THE BAND EXTENDS — the population SEVEN over THREE depths
({4130, 8222, 8234, 8294} at v₂ = 1 · {4220, 4244} at v₂ = 2
· {8432} at v₂ = 4) with a v₂ = 3 GAP as read (a flag);
era five retakes the lead (18 ≠ 15 at k = 40; the arc six → five
→ tie → six → five — ALTERNATION-CONSISTENT, the tie as
pivot; the v₂ = 2 re-tie 12 = 12 a flag); the direct-line coincidence
UNBROKEN at 160/160 across four depths. The 160-cell census QUIET 93 ·
{000, 110} 60 · ALL-BUT-101 7. MTX-A6 CLOSED AS WITNESSED —
OPEN-EDGE (seal c89431c, THE GRAMMAR ATTEMPT): THE GAP FILLS
— 4232 = ALL-BUT-101 at v₂ = 3: the band is CONTIGUOUS, and
with four new members (4262 · 4274 at v₂ = 1; 4340 at v₂ = 2)
the population is ELEVEN over FOUR depths, era-asymmetric (seven era-five vs
four era-six); THE ALTERNATION DIES at its first staked prediction — era
five persists (rich5₄₈ = 21 > 19; the arc six → five → tie
→ six → five → five: a measured pattern is exactly as local as a
received one — only the direct line predicts here); a third v₂ = 2
equality (15 = 15; the copies-behavior dominates at depth); the direct-line
coincidence UNBROKEN at 208/208. The 208-cell census QUIET 123 ·
{000, 110} 74 · ALL-BUT-101 11. The MTX-A7 hypotheses stand recorded
(§17.6), on go — the promotion-gate question stands BEYOND ripe
(Steven’s own call at a seal of his choosing).
The gate — acceptance
The six vectors (I + S1..S5) at generation 1, built and NOT frozen
(built ≠ frozen; the spec FREEZE with digests pinned + unknowing statements printed is
MTX-P0’s sealed act); the ledgers empty; the a-priori cited, never re-derived.
The DRY calibration battery — machinery verification (non-sealed)
The five text-pinned faces on the primality ground, re-read through this
experiment’s own code: 1024 / 2560 / 5120 = {000, 110} (the
persisted origin-mark bit at all three collision faces) and
2048 / 3072 = ∅. The rix BI-2 precedent: a DRY battery is not a
read. The sealed MTX-P1 ran at A1 and landed 22/22 EXACT over the full
twenty-two-cell battery — the instrument GROUNDED.
The probe cards — MTX-P0..P5
The sealed probe ladder (§6): P0 the encodings frozen · P1 the
calibration re-witness · P2 the controls (the trivial-quiet risk named) ·
P3 the prediction reads (cards: claim + exact identity + kill condition) · P4 the
folds (LINEAR vs STRUCTURAL; match or refusal) · P5 the assembly.
SEALED per attempt (the default read = the latest, MTX-A6 against
c89431c; MTX-A5 against a731aec; MTX-A4 against
f8fd71b; MTX-A3 against c47ab3b; MTX-A2 against
04320aa; MTX-A1 against 4a7dddc); UNSEALED where no seal
has committed — seal-before-read. {"live": true} re-verifies
(the policy batteries take minutes).
The external anchor
The attempt dataset to the immutable EXTERNAL record (mldata CKAN when
configured; the dummy-node fallback otherwise). The anchor dataset is created by the
operator’s first publish act; rows exist only once sealed attempts do.