skills/google/skills/design/Gen Agent Trust Hub

design

Pass

Audited by Gen Agent Trust Hub on Aug 11, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • Command Execution: The skill performs architectural tasks by executing local Python scripts and Terraform CLI commands. These operations are conducted using standard, secure subprocess methods and are necessary for the skill's intended functionality of infrastructure validation.
  • External Resource Integration: The skill fetches Terraform modules and templates from official GitHub repositories and Google Cloud Storage. These operations target verified vendor resources and are performed using authenticated CLI tools like gcloud.
  • Security-First Configuration: The skill's instructions explicitly enforce the use of GCP Secret Manager for all credentials and prohibit hardcoding sensitive data in HCL files. It also mandates version pinning for all external modules to ensure configuration stability and security.
  • Isolated Validation Workflow: Local Terraform operations (init, validate, plan) are performed in dedicated scratch directories. This isolation ensures that validation processes do not interfere with other files in the user's workspace.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 11, 2026, 08:40 AM
Security Audit — agent-trust-hub — design