Required views, context or documents are missing.
PieceID trust architecture · v1.0
Evidence risk is not counterfeit probability.
Every PieceID output must preserve the difference between an uploaded claim, a source-checked specification, technical readiness and a human specialist finding.
Allowed result states
One taxonomy across every tool and report.
The pack is usable for specialist work; no object conclusion is implied.
Scoped evidence aligns with referenced expectations.
Specific conflicts are recorded without converting them into an automatic counterfeit verdict.
Evidence cannot support a responsible conclusion.
Remote review cannot close the material question.
Identity, ownership story, purchase channel and condition statements remain attributed to the submitter.
Brand, authority or editorial sources are named and dated; marketing claims remain vendor claims.
Coverage, format and readability signals do not inspect physical construction or issue a verdict.
Reviewer identity, participation, reviewed material, limitations, disagreement and corrections must be auditable.
Credentials, covered houses/models, conflicts and escalation capacity must be documented before public “expert” labeling.
Complex watch and jewelry cases require double review or documented escalation before stronger language.
Reports are versioned. Material corrections preserve the prior version, reason, date and reviewer trail.
Cases are private by default, access-key protected, deletion-capable and retained only for the disclosed period.
Referral or commercial relationships cannot influence findings and must be disclosed.
Remote evidence cannot replace microscopy, opening a watch, material testing or in-person custody checks.