wendy-gcp-deployment
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill establishes a robust security posture for cloud infrastructure, explicitly prohibiting high-risk practices like creating service account keys or using project-wide admin roles. All architectural decisions (Cloud Run preference, VPC isolation, PSC for cross-project traffic) align with standard secure cloud design patterns.
- [DATA_EXPOSURE_AND_EXFILTRATION]: Analysis found no evidence of hardcoded credentials, sensitive file access, or suspicious network exfiltration. Network references target well-known and trusted systems (GitHub Actions, Google Cloud APIs) or the vendor's own documentation (Linear), which is consistent with the skill's purpose.
- [REMOTE_CODE_EXECUTION]: The skill does not perform remote code execution. It provides templates for Pulumi (Go) and GitHub Actions which are intended to be reviewed and executed by the user in a controlled environment. Dependencies like the Pulumi SDK and Google GitHub Actions are standard industry tools.
- [INDIRECT_PROMPT_INJECTION]: The skill interacts with user-provided requirements through an interview process. It mitigates potential risks by mandating human review of privileged bootstrap commands and requiring pre-deployment validation of existing infrastructure resources (e.g., DNS record checks) before making changes.
Audit Metadata