Credentials you can audit.
Every issued LockedIn Labs credential resolves to a public entry — the holder, the level, and the artifact trail that earned it. The cards below are specimens, not issued credentials.
- 01CREDENTIAL → ARTIFACTS
A credential resolves to the graph nodes it stands on and the artifacts submitted against them. A node with no artifact is printed unresolved, not omitted.
- 02ARTIFACT → REVIEW
Each artifact resolves to its reviews. A review records the rubric it scored, the reviewer, the date, and the verdict.
- 03REVIEW → KIND
Each review records whether it was produced by a model or by a human practitioner. The two are stored as separate values and are never merged or inferred.
- 04NO TRAIL, NO BADGE
Verification requires a live model review or an expert review. An offline heuristic check is recorded and can never mark an artifact verified.
Every entry listed below comes from the seed registry: authored demonstrations of the record format, held by no one, labelled SPECIMEN on the list and on the record. Open one and the chain above is what you read — the artifacts, the rubric each was scored against, the reviewer, and which kind of review it was.
For employers
Read a hire like a diff. An issued entry names the credential, the level, and the artifact trail behind it, with review status and provenance visible instead of collapsed into one badge.
For learners
Your work backs the credential. Graph-qualified, live-reviewed evidence can create an in-review record; activation remains a separate human decision. Formal live defenses begin only with vetted roster coverage.
For auditors
A two-year term is computed from the issue date carried on the record. Activation is a separate human decision and the schema records no date for it, so the term is never dated from one. Every entry carries its status on its face — in review, active, or expired — so currency is never something you take on trust.