Public claim register — v0.1.0 draft candidate
Status: candidate register; no row is an authorization receipt
Controlling protocol: ../../docs/research/PUBLIC_CLAIM_RECEIPT_OPERATIONS.md
Each published statement must receive a machine-readable publication-claim-receipt for exact wording and surfaces. This table records current candidates and blockers; it cannot be cited as approval.
| Candidate ID | Requested claim | Current evidence state | Current allowed treatment | Blocker / next receipt |
|---|---|---|---|---|
| CAND-STATUS-001 | Ontology v0.1.0 contains 10 domains and 55 capability families | Reproducible from released ontology and depth validator | May be described as released source structure when bound to final manifest | Claims reviewer, guarantor, and release authority receipt |
| CAND-METHOD-001 | Synthetic scoring demonstration reproduces its expected output | Executable synthetic fixture and runner exist | Must retain synthetic, experimental, and not a human validation result boundaries |
Final environment/output digest and claim approvals |
| CAND-STATUS-002 | Official human scoring is disabled | Repository controls and status documents state disabled | May be described as current product state when build/live evidence matches | Final deployment evidence and claim approvals |
| CAND-COUNT-001 | At least 500 AI practitioners were interviewed | Unverified requestor assertion; zero records admitted | Only the exact held wording in the aggregate claim receipt | Underlying records, eligibility/count reproduction, full study gates and approvals |
| CAND-COUNT-002 | At least 50 executives were interviewed | Unverified requestor assertion; zero records admitted | Only the exact held wording in the aggregate claim receipt | Underlying records, eligibility/count reproduction, full study gates and approvals |
| CAND-RATE-001 | Fewer than 5% would pass | No study, authorized score, sample, threshold, or result | Prohibited | Validity, standard setting, representative use evidence, study and claim authority |
| CAND-REVIEW-001 | Independently reviewed / peer reviewed | No named completed review | Prohibited | Named appointment, independence assessment, exact scope, signed disposition |
| CAND-ARCHIVE-001 | Citable archived release / DOI | No verified deposit or registered identifier | Prohibited | Approved licenses/authorship, archive deposit, DOI/PID and independent verification |
| CAND-VALIDITY-001 | Predicts field performance or hiring success | No outcome-linkage study | Prohibited | Prospective criterion evidence, replication, fairness/accessibility and legal/use review |
Do not remove a blocked row to make a release appear cleaner. Supersede it with a receipt and preserve history.