vertex-ai-gemini
Pass
Audited by Gen Agent Trust Hub on Aug 12, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill does not contain instructions that attempt to override agent behavior or bypass safety guidelines. The instructional content is standard technical documentation for API integration.- [DATA_EXPOSURE_AND_EXFILTRATION]: No sensitive data or hardcoded credentials were found. The skill demonstrates best practices by using environment variables and Service Account placeholders for authentication.- [EXTERNAL_DOWNLOADS]: The skill references official Google Cloud SDKs including google-genai and google-cloud-aiplatform. These are well-known packages from a trusted organization.- [INDIRECT_PROMPT_INJECTION]: The skill includes code for processing external data from Cloud Storage (gs://) and local files. 1. Ingestion points: Data is ingested via Part.from_uri and Part.from_data in SKILL.md. 2. Boundary markers: None provided in the simplified code examples. 3. Capability inventory: The skill uses network operations to communicate with Google Cloud Vertex AI endpoints. 4. Sanitization: No explicit sanitization of input content is shown in the provided snippets.
Audit Metadata