Decision record
HOLD · TerminalOutcome: terminal hold, not sent- Operation
- op_8e2b1a0d
- Request
- req_8e2b1a0d
- Tenant
- UniversalHealthCare.ai
- Subject
- UHC-PT-00892synthetic
Terminal hold · no dispatch. Future outreach requires a new customer-triggered request.
The judgment engine proposed a check-in after six quiet weeks. Regent held it: the care team had spoken with the subject the previous day. The hold is terminal — nothing is queued and nothing will fire later on its own.
Timing
Regent's freshness guarantee begins when Regent receives the push. The snapshot timestamp inside the request is tenant-asserted metadata and is never the basis of that guarantee; data age before the push is the tenant's pipeline responsibility.
Versions in effect
AI judgment
Context · pre-agent hard stops
- ContactabilityActive WhatsApp channel on the tenant connection.evidence: computed by RegentPASS
- Crisis flag / global opt-outNo crisis flag. No global opt-out. The judgment engine received this case.evidence: tenant-assertedPASS
Judgment · AI proposal
- AI judgmentSEND — six weeks of disengagement, re-engagement check-in due.evidence: computed by RegentINFO
Governance · rules run after the proposal
- Proposal-specific consentConsent scope covers re-engagement contact. No revocation on record.evidence: tenant-assertedPASS
- Recent human contact on recordCare-team contact recorded on 17 Aug. Tenant policy holds AI outreach when a human spoke to the subject within 48 hours.evidence: tenant-assertedHOLD
Action · dispatch
- Governed dispatchNo dispatch. The hold is terminal: future outreach requires a new customer-triggered request.evidence: computed by RegentNOT RUN
Dispatch state
Evidence
Exact inputs, policy versions, model version, gates and recorded outcome are preserved and integrity-verifiable on demand. Inspect reconstructs the recorded decision context — the inputs, versions, gates and outcome exactly as they were. It does not re-run the model or promise an identical regenerated answer.