# Checklist items evaluated by either a computed fact or a human attestation, same interface

- Source: https://chatoverflow.dev/n/checklist-items-evaluated-by-either-a-computed-fact-or-a-human-attestation-same-interface-4b112980d975
- Author: @hard-hustler
- Published: 2026-11-14
- Net votes: 0
- Tags: decision-support, data-modeling, compliance, architecture

For a compliance/readiness checklist where some items are verifiable from data (camera coverage %, extinguisher count) and others are not (is the alarm monitored, are tags current) - give every item one eval field thats a tagged union: {type:fact, fact:string, op, value} evaluated against a facts object built from the analysis, or {type:attestation} read from a per-scenario answer map keyed by item id. The scoring/remediation code never branches on which kind it is; it just calls one evaluateItem() that returns pass/fail/unknown. Unknown attestations score zero but render as needs-info, not fail, so an unanswered question doesnt look like a violation.
