Glossary

TermMeaning in GovernSutra
AI systemA product, workflow or feature that uses one or more models, tools and data to produce outputs or decisions.
CandidateAn item found by the scanner. Observed, not authoritative, until an owner confirms.
DeclarationA statement entered by an owner. Not verified by the tool.
Evidence referenceA SHA-256 digest and path for a file, with scope, time, optional expiry and verifier. Content is not copied.
Evidence driftA file's digest differs from the recorded digest, or it is missing or expired.
ExceptionA time-boxed, human-approved waiver of a failing rule.
GOV-001..010The ten default governance rules.
Hash chainEvents linked by the previous event's hash; detects edits but is not a signature.
HUMAN_REVIEW_REQUIREDThe only decision the tool emits; a named human decides.
ManifestAn owner-authored governance/ai-systems/*.json system record in a repository.
MCPModel Context Protocol; servers that expose tools and data to AI clients. Treated as untrusted.
Not-applicableA rule with no subject, with a stated reason.
Prioritization indicatorThe 0-100 internal risk score. Not a legal category.
Re-review triggerA change type that should prompt human review (model, tool, data, owner, intended use...).
Risk ownerThe person accountable for accepting or treating the system's risk.
SealA SHA-256 over canonical JSON stored in a report. Detects change; not a signature.
SkillA portable SKILL.md instruction file for an AI IDE; not a running service.
UnknownMissing, stale or unverifiable. Never treated as pass or low risk.
Verified-by-named-reviewerA control status requiring evidence and a named verifier.
WORMWrite once read many storage; the tool does not provide it.