cloud-run-alert-configuration

Pass

Audited by Gen Agent Trust Hub on Sep 21, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • Command Execution: The skill utilizes the gcloud CLI for discovering Cloud Run services, jobs, and worker pools. This is a standard administrative function necessary for its primary purpose of configuring monitoring for these resources. The skill includes specific rules to skip these commands if the resource name is provided in the prompt, further limiting tool usage.
  • Local File Access: The skill scans local .tf files within the immediate working directory to identify existing Cloud Run resource definitions. This is an expected behavior for a tool designed to assist with Terraform configuration management.
  • Indirect Prompt Injection Surface: The skill ingests data from user prompts and local Terraform files to parameterize alert policies. While these ingestion points represent a standard surface for AI agents, the skill implements 'Fast Path' logic and directory bounds that significantly restrict the scope of data processed, minimizing potential exposure to untrusted content.
  • Metadata and Resource Tracking: The skill mandates the inclusion of a specific user_labels block in all generated Terraform resources (created-with-google-skill). This practice enhances auditability by clearly identifying infrastructure components created or managed by the AI agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 21, 2026, 03:01 PM
Security Audit — agent-trust-hub — cloud-run-alert-configuration