Secoya One has undergone extensive synthetic testing of independent verification, cryptographic integrity, signing authority, timestamping and historical trust continuity.
The programme has also demonstrated the rejection of cryptographically valid evidence presented under an unauthorised signing identity.
The next phase is institutional deployment and assurance: integrating customer-controlled HSM/KMS infrastructure, strengthening operational key and policy lifecycle controls, completing RC1 conformance testing and evaluating Secoya within real institutional workflows.
Secoya preserves independently verifiable evidence of recorded decisions and actions. The assessment of underlying business outcomes remains with the institution and its authorised reviewers.
The current findings represent project-conducted technical validation, not independent certification or regulatory endorsement.
Methodology, tested boundaries and the assurance work that accompanies institutional deployment.
The programme summary draws on project-operator-supplied findings. Sandbox execution is operator-reported and not independently attested; this website is not an independent audit. Local implementation checks, historical Sandbox records and later synthetic demonstrations have distinct scopes. A local pass does not establish Sandbox execution, and later work does not automatically close earlier unresolved items.
The supplied programme reports SDK-independent reproduction of selected verification behaviours. Earlier historical lending runs used the Secoya SDK. Selected parser, boundary and canonicalisation results do not establish full RC1 verifier conformance, which remains part of the next phase.
The counterfeit package was internally consistent under its own supplied trust material. It was rejected under independently pinned external policy at the signer gate. Lifecycle and timestamp authority controls were tested separately; the integrated rejection does not establish that every downstream gate ran.
A separately signed assessment can omit a required input while remaining cryptographically valid. Completeness requires an expected-input or policy reference distinct from signature validity. Neither check establishes source truth or universal detection of missing events.
RFC 3161 testing used a local test timestamp authority. Lifecycle demonstrations used controlled or hypothetical historical metadata, not independently established institutional history. These results do not establish a production timestamp service or retrospectively upgrade earlier signer-declared timestamps.
Reported local HTTP, webhook and sidecar scenarios exercised invocation alongside existing workflows. Consumer Duty HTTP notebook observations are distinct from the standalone issuance-adapter reproduction. Institutional HSM/KMS integration, authenticated policy rotation, rollback protection, authenticated durable delivery, retries, idempotency and durable recovery require further validation.
Historical lending evidence preserves recorded outcomes; it does not independently reproduce the original lender’s decision process. Synthetic technical findings do not establish live-customer outcomes, measured return on investment, fairness or Consumer Duty compliance. Substantive assessment remains with the institution and its authorised reviewers.
Independent external review of underlying evidence and explicit closure of unresolved items remain part of the assurance programme. Detailed Sandbox counts, artifacts and downloads are not published here; disclosure requires separate approval. Deployment and governance controls must be evaluated within the agreed institutional workflow.