gcp-architect
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill enforces security-first principles for GCP architecture. It provides explicit instructions to use Google Secret Manager for sensitive credentials and avoid environment variables or hardcoding, which is a key security best practice.
- [SAFE]: The provided Cloud Run deployment template (resources/cloud_run_deploy.yaml) correctly implements secret referencing via secretKeyRef and follows networking best practices by using a custom VPC and Direct VPC Egress.
- [INDIRECT_PROMPT_INJECTION]: The skill acts on architectural requirements provided by the user. While the provided files do not contain code-execution tools, the agent's guidance could be influenced by malicious prompts embedded in design requirements. (1) Ingestion points: User queries regarding system design and infrastructure requirements. (2) Boundary markers: Absent within the skill instructions. (3) Capability inventory: No executable scripts or tool calls are defined within the skill files or templates. (4) Sanitization: Not present; the skill consists of static instructional guidelines and configuration templates.
Audit Metadata