Public ledger

No casework has cleared the evidence gate yet.

Existing portfolio summaries are not automatically casework. A named engagement, implementation description, or favorable number does not earn admission without the evidence needed to reconstruct the reasoning chain.

0 admitted cases

Admission contract

A case must preserve seven separate records.

  1. 01

    Observed condition

    What the system demonstrably returned before intervention, stated without explanation or attribution.

  2. 02

    Evidence

    The raw response, export, DOM, request log, configuration, timestamp, hash, or other artifact that establishes the condition.

  3. 03

    Decision

    Why the selected intervention followed from the evidence, including alternatives considered and the rule used.

  4. 04

    Intervention

    The exact release, configuration, content, routing, or system change, with scope and deployment identity.

  5. 05

    Verification

    The same or preregistered test repeated after the change to establish whether the targeted condition still exists.

  6. 06

    Result

    Only the conclusion supported by the evidence. A verified technical correction is not automatically a traffic or ranking effect.

  7. 07

    Limitations

    Unavailable evidence, alternative explanations, unmeasured outcomes, confounding changes, and conclusions the case cannot support.

Claim boundary

Verification is not the result.

Verifiable technical statement

“The canonical disagreement was eliminated.”

This can be supported by repeating the same extraction across HTML, JSON-LD, sitemap, and response behavior after release.

Required evidence: before/after artifacts, frozen normalization rule, release identity, and repeated test output.

Separate outcome claim

“Organic traffic improved because the disagreement was fixed.”

This requires outcome data and a design capable of supporting attribution. The technical verification alone cannot establish it.

Additional evidence: pre/post time series, unit of analysis, comparison or counterfactual, exclusions, uncertainty, and confound review.

Reusable case record

Minimum publication package

case_id
organization_or_subject
scope
observed_condition
baseline_artifacts[]
decision
alternatives_considered[]
intervention
release_id
deployed_at
verification_procedure
verification_artifacts[]
result
outcome_evidence[]
limitations[]
methods_used[]
approved_by
approved_at

A field may be unavailable, but it may not be silently omitted or reconstructed from the final claim.

Methods used

Casework inherits the public methodology.

Decision TracesSeparates observation, evidence, interpretation, decision, action, and verification. Silent Hydration InvestigationDemonstrates one reproducible technical evidence chain. Search System DiagnosticProduces bounded public observations without a synthetic score.