What a Supervisor Can Verify

Every measurement on this site is auditable by qualified scrutineers.

6
Verifiable artefacts available under NDA
SHA256
Cryptographic seal at every pipeline stage
8/8
Pre-registered predictions held

A supervisor who acts on a Meridian finding requires the ability to verify the finding. The substrate exposes six artefacts that allow that verification under non-disclosure agreement. The cryptographic seal fingerprints the entire substrate at SHA256 e5250de8e9de07d6, permitting bit-identical reconstruction of any reported finding. The sealed snapshot manifest freezes the population at v13.1.0-production-day-zero-full as of 29 April 2026. The capture-replay validation cycle on an anonymised G-SIB documents byte-identical input layer behaviour across four observations and bounds the model variance at the extraction layer. The per-edge evidence tier permits a supervisor to filter any finding to observed evidence only, removing inference from the analysis. The per-agent unknown state distinguishes "the institution has no halt mechanism" from "the substrate has not established whether the institution has a halt mechanism." The pre-registered prediction log documents the eight predictions that were stated before the substrate was sealed and the eight that subsequently held. A measurement that cannot be audited cannot be acted upon.

Figure 1 · Verifiable Artefacts
SIX VERIFIABLE ARTEFACTS · AVAILABLE TO QUALIFIED SCRUTINEERS UNDER NDA Substrate cryptographic seal SHA256 e5250de8e9de07d6 Reproducible to byte Sealed snapshot manifest v13.1.0-production-day-zero-full Frozen 29 April 2026 Capture-replay validation Anonymised G-SIB · 4 observations Byte-identical input layer Per-edge evidence tier Observed / Derived / Inferred Source attributable Per-agent unknown state ODIU four-state taxonomy No silent compromise Pre-registered prediction log 8/8 confirmed at v13.1.0 sealing Auditable A measurement that cannot be audited cannot be acted upon. MAR®500.com

Six discrete artefacts ground the substrate's verifiability claim. Each is reproducible by a qualified scrutineer under non-disclosure agreement, with the methodology fully documented in The Stationary Sea (Part 1) on Zenodo. The cryptographic seal anchors the substrate to a specific point in time and a specific scanner version. The capture-replay validation cycle proves byte-identical retrieval-layer behaviour. The four-state evidence taxonomy preserves interpretive discretion at the point of analysis: a supervisor may compute any finding on the full substrate, on observed and derived edges only, or on observed edges alone, producing the three views of any composite at the substrate level. The companion paper Stationary Sea (Part 2: Rating Methodology) seals the rating layer's specific calibrations via cryptographic manifest commitments, available for verification by qualified scrutineers.

Source · Stationary Sea Part 1 (Zenodo) · Substrate v13.1.0 sealed 29 April 2026 · [email protected] for verification access Methodology · Stationary Sea Part 1 (Zenodo)
543 institutions · 12 sectors · 95,876 agents · 636,854 governance edges · substrate v13.1.0
Methodology grounded in The Stationary Sea (Part 1) on Zenodo. Companion: Stationary Sea Part 2 (Rating Methodology) sealed via cryptographic manifest fingerprint.