skills/google/skills/skill-registry/Gen Agent Trust Hub

skill-registry

Pass

Audited by Gen Agent Trust Hub on May 14, 2026

Risk Level: SAFE
Full Analysis
  • Official Google Cloud Integration: The skill uses standard Google Cloud authentication patterns (gcloud auth application-default login) and official libraries (google-auth, requests) to interact with the Skill Registry.
  • Network Communications: All network operations in scripts/skill_registry_ops.py are targeted at *-aiplatform.googleapis.com. These are official Google API endpoints used for managing enterprise agent resources.
  • Data Management and Uploads: The upload and update functions in scripts/skill_registry_ops.py handle local files and folders by zipping and Base64-encoding them for transmission to the registry. This is a standard method for deploying skill packages to the platform.
  • Environment Ingestion Surface: The skill reads configuration from GCP_PROJECT_ID and GCP_LOCATION environment variables. This is a standard practice for configuring cloud-based automation tools.
  • Indirect Data Processing: As a registry management tool, the skill fetches and displays data (such as skill names and descriptions) from an external source. While this represents a surface where external instructions could be viewed, the skill performs read-only operations for discovery and uses standard JSON formatting for output.
Audit Metadata
Risk Level
SAFE
Analyzed
May 14, 2026, 11:16 PM
Security Audit — agent-trust-hub — skill-registry