google-cloud-solution-agentic-ai-borderless-data-lakehouse

Pass

Audited by Gen Agent Trust Hub on Sep 18, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill retrieves documentation and technical guides from external domains including docs.cloud.google.com, codelabs.developers.google.com, and registry.terraform.io. These references target official Google Cloud and HashiCorp platforms for architectural grounding and implementation details.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external content to inform its design and implementation outputs, creating a surface for indirect prompt injection.
  • Ingestion points: External resource retrieval occurs in SKILL.md under Phase 2 Step 1, Phase 3 Step 1, and Phase 4 Step 1.
  • Boundary markers: None identified; there are no explicit instructions to the model to ignore potential injection attempts within the retrieved documentation.
  • Capability inventory: The skill possesses the ability to generate Terraform files and shell scripts (using curl and gcloud) based on the ingested data.
  • Sanitization: No validation or filtering mechanisms for the external data are described in the instructions.
  • [DYNAMIC_EXECUTION]: The workflow involves generating Infrastructure as Code (Terraform) and validation scripts (shell commands) to provision resources and verify connectivity. While these scripts are presented for user review and manual execution, they represent dynamically generated logic derived from external technical resources and user input.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 18, 2026, 05:40 AM
Security Audit — agent-trust-hub — google-cloud-solution-agentic-ai-borderless-data-lakehouse