gcp-secret-manager
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides implementation guidelines for Google Cloud Secret Manager using official gcloud CLI commands and verified SDKs for Python, Go, and Node.js.
- [SAFE]: Demonstrates secure deployment patterns, specifically utilizing Kubernetes Secrets Store CSI Driver and GKE Workload Identity to avoid the use of long-lived service account keys.
- [INDIRECT_PROMPT_INJECTION]: The skill outlines the workflow for ingesting secret data into application environments, which is an inherent part of secret management.
- Ingestion points: Secret payloads retrieved from the GCP Secret Manager service via CLI or SDK calls (SKILL.md).
- Boundary markers: Absent; the skill focuses on management and access rather than the content validation of the secrets themselves.
- Capability inventory: Writing secret data to local files (e.g., /private/new-secret-output) and mounting secrets as environment variables or volumes in GKE (SKILL.md).
- Sanitization: Examples utilize standard decoding and JSON parsing; the skill assumes secrets are managed by authorized personnel.
Audit Metadata