agent-platform-prompt-management
Pass
Audited by Gen Agent Trust Hub on Nov 3, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- Resource Management via SDK: The skill utilizes the
google-cloud-aiplatformandgoogle-genailibraries to perform operations like creating, listing, and deleting prompts. These are standard vendor-provided libraries for interacting with Google Cloud services. - Safety Confirmation Tiers: The skill implements a structured safety model (Tiers R, M, and D) to prevent accidental data loss. Destructive actions like deletion require explicit, high-friction user confirmation before execution, which is a security best practice for autonomous agents.
- Dependency Installation: The skill probes the environment for necessary Python packages and installs them if missing. While this involves downloading external code, it targets official packages (
google-cloud-aiplatform,google-genai) from the standard registry (PyPI). - Usage Attribution: It includes instructions for attributing API calls to the specific skill using environment variables and User-Agent headers, which is standard for usage monitoring and does not involve exfiltration of user data.
- Environment Setup: The instructions guide the user through standard Google Cloud authentication (
gcloud auth login) to ensure operations are performed within the user's authorized security context.
Audit Metadata