Anchor · Validation

Satoshium Anchor · Validation

Validation determines whether an Integrity Reference satisfies Anchor's structural and institutional requirements before it is allowed to proceed toward public authority.

Validation evaluates the Anchor record. Verification evaluates the integrity evidence. Publication requires both.

Validation Principle

Validate the institution before publishing the record.

Validation vs. Verification

Validation → Does the Integrity Reference satisfy Anchor's requirements?

Verification → Does the Reviewed Representation match the preserved integrity evidence?

A record may pass one and fail the other. Neither alone is sufficient for publication.

PASS

All applicable required Validation rules have been satisfied for the reviewed Anchor Version.

Overall Outcome

FAIL

One or more applicable required Validation rules have not been satisfied.

Overall Outcome

Rule-Level Review

Each applicable Validation rule produces PASS or FAIL. Non-applicable rules may be explicitly recorded as NOT APPLICABLE where needed.

Rule Behavior

Publication Evidence

Validation PASS becomes one of the required inputs to the formal Publication Gate.

Gate Input

Production Sequence

Integrity Reference constructed

Anchor Identifier assigned

Anchor Version assigned

Formal Validation

Initial Verification

Canonical HTML / JSON consistency

Publication Gate

Published Integrity Reference

Identifier Assignment Resolved

Formal Validation must evaluate the actual production candidate, and the Base Schema requires anchor_identifier.

Anchor Identifier assignment → before formal Validation

Assignment does not create public authority.

Anchor Identifier assigned ≠ Published ≠ Publication Gate APPROVED

Once assigned to a production candidate, an Anchor Identifier should not be recycled for a different Integrity Reference.

Architecture Decision

Validation Rule Families

Identity
Source
Representation
Integrity Method
Provenance
Relationships
Record State
Versioning
Corrections
Verification Readiness
Publication Readiness
Human / Machine Consistency

Initial Rule Set

The first governed Anchor Validation Rule Set contains:

VAL-001 through VAL-042

The detailed definitions live in anchor-validation-rules.md.

1.0-draft

Identity & Source

Validates Anchor identity, Version, Schema Version, Source Institution, Source-System Identifier, Source Artifact type, and authority coherence.

Representation

Validates Representation Type, Representation Boundary, canonicalization, reconstructability, and integrity-subject coherence.

Integrity Evidence

Validates method presence, method completeness, Integrity Value or proof material, and permitted method / algorithm use.

Provenance

Validates Source, representation, and generation provenance and whether the integrity process can later be reproduced or historically validated.

Relationships & State

Validates Source linkage, relationship coherence, supersession lineage, separate record-state dimensions, and Lifecycle values.

Versioning & Corrections

Validates subject continuity, Version history, new-identity requirements, Correction lineage, and Source-change boundaries.

Schema Validation Is Not Enough

JSON Schema can validate structure without proving institutional meaning.

Schema valid

Anchor valid

For example, two strings may satisfy JSON Schema while still improperly using the Source-System Identifier as the Anchor Identifier. Anchor Validation must catch institutional errors that Schema alone cannot.

Lifecycle Validation

Validation now enforces the frozen Lifecycle State vocabulary:

draft
active
superseded
withdrawn
archived

It also checks allowed transitions and rejects silent reverse transitions.

Versioning Validation

same integrity subject → Anchor Version

new integrity subject → new Integrity Reference

Validation must reject attempts to disguise a materially different integrity subject as merely a later Version.

Correction Validation

If a Version contains a Correction, Validation checks that prior state, corrected state, reason, Version lineage, and subject-continuity decision remain reconstructable.

Correct forward. Preserve backward.

Publication-Readiness Validation

Before the Publication Gate, Validation confirms the presence of:

Validation PASS
Initial Verification completed
canonical HTML
canonical JSON
human / machine consistency
appropriate Lifecycle State
no blocking Correction
no governance hold

Human / Machine Consistency

Canonical HTML

Canonical JSON

Material facts must agree before Publication Gate approval.

Validation Record

A formal Validation execution should eventually preserve:

Anchor Identifier
Anchor Version
Schema Version
Validation Rule Set Version
validated_at
individual rule results
overall PASS / FAIL

Whether Validation Records receive independent identifiers remains unfrozen.

First Production Test

The first production Integrity Reference should test the Validation architecture itself.

Which rules were necessary?
Which were redundant?
Which were ambiguous?
Which should become machine-enforced?
Which requirements were missing?

Production should refine the rule set rather than merely prove compliance with assumptions we wrote in advance.


A record should satisfy the institution before the institution stands behind the record.