blog(cook): the context you are paying to re-send — taken, the next morning: the post's own break-it broke the meter by 40× (115 calls · 19.7 M tokens on a 2.0 MB transcript; the unit of re-send is the call, not the turn), the meter now reads usage rows, the cog and its first pegs, the three regimes on one card, effectiveness 0.59, and C91 as the row that ends the orchestrator; validate-mdx green, mdx-cook structure green (S12 ratchet red is the floor another post set — named, not sanded)
This is a live, recomputable drift receipt — the on-chip proof of where this commit's work landed (in its authorized domain, out of it, or unplaced), sealed with a signature anyone can verify. It is the same receipt shipped in the on-commit email, hosted here so the claim is checkable on the open web. The full receipt (attestation, density panel, full commit message) publishes at /commit/4d8ef1803/ shortly after the commit lands.
This is the proven asset: the drift read was produced by a real on-chip walk on sensor: metal, sealed with an ed25519 signature over the canonical body, and bound to the exact lattice by gridHash. Anyone can recompute it and check the signature — no trust in us required.
These are measurements, not a validated risk score. In a pre-registered predictive test the drift metric did not predict real defects (AUC ≈ 0.50, at chance) and the σ is pinned — so nothing here is a probability of loss and nothing here is a price. It is honest telemetry we publish while the instrument is under calibration.
Δ weighted bleed · Δ-spread max king-move leak · Γ traction slope · 𝒱 reef volatility · Θ efficiency decay (null → pending; no per-commit token meter). Definitions are carried in the signed signal's notes.
Sufficient for: where it landed, re-runnable from the commit: same bytes, same hash.
NOT sufficient for: whether the work was good, which is undecidable (Rice, 1953); nothing here claims it. Rice (1953): no program can decide, in general, whether another program is correct.