agent-platform-model-registry
Pass
Audited by Gen Agent Trust Hub on Jun 13, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the
gcloud ai modelscommand-line interface to perform model registry management tasks. This includes listing existing models, retrieving metadata, and performing administrative actions like uploading new versions or deleting models. - [SAFE]: The skill implements a robust 'Safety & Confirmation Tiers' framework. This system classifies actions into read-only, mutating, and destructive categories, mandating that the agent present literal command strings for user review and wait for explicit confirmation in a subsequent interaction turn before execution. These measures effectively mitigate risks of accidental or unauthorized resource modification.
Audit Metadata