Kill Switches

Two switches arm at this lock on the author’s word, and one of AP30’s is paid.

KS-HOLD.1 — The closed form

Claim. m_p/mₑ = 1836 + 21α·(1 − 1/(84π)) + 21α·r/(1 − r), r = 16α/1836: D_closed = 1836.152673444951, the third-order coefficient c₃ = +21·16²/1836² = +1.595 × 10⁻³ with its sign stated in advance, all higher orders summed, with no free parameter. Supersedes the corpus’s registered D of 2 August 2026 by reference to its digest, beside it and not in its place.

Test. Dies at 3σ once the proton-to-electron mass ratio is measured to 3.4 parts per trillion with the present central value unchanged — about five times today’s precision — and at 5σ at 2.1 parts per trillion; the third-order term itself is resolved at about fifty times today’s. Fires in either direction. No negative term exists in the chain with which to repair a growing negative residual, and none will be offered.

Status. LIVE — EMPIRICAL; the kill within the horizon the corpus’s register gives as three to five years.

Recovery. The sixteen and its reading (§3) stand on the second-order data; the chain’s rule (§4) retires; AP30’s formula returns to its truncated form with KS-30.3 re-opened.

KS-HOLD.2 — The reading

Claim. A second-order support count in the corpus is its object through the four dimensional expressions: the proton’s channels, 21 × 4; the neutron’s pair, 2 × 4; the hold’s conditions, 4 × 4.

Test. Fires if any second-order count in the family — at the neutron, at G, or at a site not yet written — is found to require its object squared, or its object through three dimensions, or any count not of this form.

Status. LIVE — STRUCTURAL.

Recovery. The sixteen stands on the measurement as a carried count; its reading retires to STATED; KS-30.3 re-opens on its first limb.

KS-HOLD.3 — The base of the chain

Claim. The second holder holds against the repair’s bare coupling, 21α, and not against the repair net of its returned leak, 21α·(1 − 1/(84π)): the chain’s tail is 21α·r/(1 − r), which is AP30’s own second-order term continued. In the author’s words the hold acts on the expression event — “expression of structure pays” — and the returned leak is not an expression.

Test. The two bases predict 1836.152673445 and 1836.152673408, 3.7 × 10⁻⁸ apart. CODATA 2022 sits between them, 0.59σ below the bare value and 0.56σ above the leaked one; it decides nothing yet. At ten times today’s precision, 1.7 parts per trillion, the two are twelve sigma apart; at 3 parts per trillion, about seven. Fires if the measurement settles on the leaked value, or on neither.

Status. LIVE — EMPIRICAL; decided by the same measurements that test KS-HOLD.1.

Recovery. The sixteen and its reading stand; the chain’s rule stands with the leaked base substituted, and D_closed is re-registered at the leaked value by dated note, never by edit.

KS-30.3 (AP30) — paid

Claim, as registered (AP30 §7, verbatim). “The perturbative series converges, with each successive order smaller by roughly a factor of α. The residual after all orders is ε operating recursively — the break maintaining itself at every scale — and must be finite.” Test, as registered. “Compute the explicit O(α²), O(α³) forms and show they misbehave — do not shrink, or do not close.” Status. PAID at the structural register on the author’s rule: the chain is ruled in §4, the explicit forms of every order follow from it and are tabulated, the ratio is r = 16α/1836 — smaller than α by 16/1836 — and the series converges by construction. The payment is a ruling and its consequence, not a computation from the mechanism, and is graded so in §0.4.

Why three

A switch on the sixteen’s digit would be a switch on AP30’s data, which already pins it. The closed form and its sign are KS-HOLD.1; the grammar is KS-HOLD.2; the chain’s base is KS-HOLD.3, which the same measurements decide. Nothing else in the paper is a claim with an edge.