CMM.X State Infrastructure PUBLIC
LIVEVERIFY
CLAIM BOUNDARY predictive_claim = NONE Diagnostic + verification + history layer. predictive_claim = NONE. No edge, no alpha, no direction, no trading signal. Under prospective test.

Ratified state

No axis is ratified as a carrier. Four independent measurement rounds failed to reject the null hypothesis — most recently PHI@1m at 4000 draws with p ≈ 0.65 against a correction threshold of 0.005116, measured 2026-08-22. An earlier finding (p = 0.00995, measured 2026-08-22) turned out to be an artefact of a moving data window and has been withdrawn.

KEEP / DEMOTE answers a different question: redundancy between axes (max|corr|), not carriage. An axis can be distinct and still carry nothing. KEEP is not proof.

AUTHORITATIVE_AXES
0/5
AUTHORITATIVE_CELLS
0/40
ACTION_GATE
0
FORWARD_SOURCE_OBSERVATIONS
0

CELLS DERIVED (keine Achse traegt -> keine Zelle kann autoritativ sein) · ACTION_GATE MEASURED · FORWARD MEASURED ueber 6 Erfassungssaetze

AUTHORITATIVE_AXES measured 2026-09-11 · UNDETERMINED_SUPERSEDED_METRIC

Live values on this page0 Reasoncopies go stale silently Numbers are on/axes & /verify

A record of state. Not a prediction.

CMM.X reads market structure into bounded, non-directional measures, seals every state into an append-only hash chain and publishes the evidence so it can be re-derived independently. It does not say where a price goes, and does not claim it could.

What is measured

Position in windowWhere the last close sits between the high and low of its window — separately per time scale.
DiscriminabilityWhether an axis measures something no other already measures, and whether it moves at all.
MagnitudeThe amplitude of the structural state. Large means moved, not up.
IntegrityWhether every block hangs cryptographically on the previous one and nothing was altered afterwards.
ReproductionWhether a value recomputed from its inputs yields the same result.

What does not follow

DirectionNO
Entry or exitNO
RecommendationNO
Proven advantageNO

These are not cautionary caveats but measurement results. Four independent rounds tested whether an axis beats a null hypothesis. None showed it. A fifth finding that appeared to show it turned out to be an artefact of a moving data window and has been withdrawn.

How errors are handled

Wrong findingsstay in place and are marked as withdrawn — never silently replaced.
Sealed artefactsare never rewritten; status changes go through a separate register.
Missing dataappears as UNDETERMINED — never as 0 and never as a substitute value.
Known defectsare stated on the page where they take effect, not in a footnote.

Where the numbers are

Front pagethe present structural state, by the minute
System statewhether the apparatus runs, and what follows from that
Axesdiscriminability and present magnitudes
Verifychain, signature, reproduction ratio
Pre-registrationsealed before, counted after
Assetsthe same structural language across several markets
next question Where does it stand right now? The boundary is drawn. Behind it lies a measured present state, updated every minute.