gcp-agent-eval-engine-runner

Pass

Audited by Gen Agent Trust Hub on Mar 11, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The script imports vertexai, google-genai, and httpx. These are well-known libraries and official SDKs for interacting with Google Cloud services and performing asynchronous HTTP requests.
  • [COMMAND_EXECUTION]: The script utilizes httpx to communicate with external agent endpoints and Vertex AI APIs. This functionality is central to the skill's purpose of automating agent evaluation pipelines.
  • [DATA_EXFILTRATION]: The script accesses the PROJECT_ID environment variable to configure the Vertex AI client. It transmits prompt data and agent responses to Google Cloud's evaluation service as part of the intended evaluation workflow.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 11, 2026, 11:49 AM