gcp-cloud-run

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • [General] (SAFE): The skill contains standard boilerplate and configuration for GCP Cloud Run and Functions. No malicious patterns, obfuscation, or data exfiltration attempts were found.
  • [Privilege Escalation] (SAFE): The provided Dockerfile follows best practices by using node:20-slim and switching to a non-root USER node for execution.
  • [External Downloads] (SAFE): All external references point to trusted Google Container Registry images (gcr.io) and official Cloud SDK tools.
  • [Indirect Prompt Injection] (LOW): The skill defines functions that process external input (e.g., HTTP request parameters, Pub/Sub messages). While this creates an attack surface for the resulting application, it is inherent to the service's purpose, and the provided templates use standard parsing methods.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:38 PM