gcp-cloud-architect
Pass
Audited by Gen Agent Trust Hub on Sep 4, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill's primary function is to assist in GCP infrastructure design. A thorough audit of the included Python scripts and documentation confirms they are legitimate tools with no malicious intent. The scripts perform standard data processing and string generation for configuration purposes.
- [COMMAND_EXECUTION]: The skill generates deployment scripts (gcloud and Terraform) as part of its deployment management features. These commands are generated based on architectural patterns and are intended for the user to review and execute, representing standard DevOps utility rather than a security risk.
- [CREDENTIALS_UNSAFE]: No hardcoded credentials or API keys were found. The skill explicitly promotes the use of Google Cloud Secret Manager for handling sensitive information in its reference documentation.
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided inputs via requirements and inventory JSON files to generate architecture designs and cost reports. While this constitutes a data ingestion surface, the processing is performed by static Python logic that does not interpret external content as commands for the AI agent.
Audit Metadata