HCS-25: AI Trust Score Methodology
HCS-25 defines how agent trust scores are computed from normalized external signals, persisted in registry metadata, and exposed through score-compatible adapters.
In progress
The signals and adapters catalogs for this standard are being assembled. Refer to the canonical specification for the current definition.
- Specification: HCS-25 in hiero-consensus-specifications
- SDK support: the
@hashgraphonline/standards-sdkHCS-25 module provides metric evaluation and validation schemas.