azure-validate
Pass
Audited by Gen Agent Trust Hub on May 3, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill and its reference files contain no malicious patterns or security vulnerabilities. All operations are aligned with legitimate cloud development tasks.
- [PROMPT_INJECTION]: The skill uses authoritative guidance markers (e.g., 'AUTHORITATIVE GUIDANCE', 'supersedes prior training') to ensure adherence to specific deployment procedures. This is acceptable in the context of professional development workflows.
- [COMMAND_EXECUTION]: The skill invokes official tools such as the Azure CLI, AZD, Terraform, and Docker. These are used strictly for validating project readiness, building resources, and checking authentication status.
- [PROMPT_INJECTION]: The skill presents an attack surface for indirect prompt injection by reading project files. 1. Ingestion points: Reads 'azure.yaml', '.azure/deployment-plan.md', and infrastructure templates (Bicep/Terraform). 2. Boundary markers: Absent. 3. Capability inventory: Executes subprocesses for 'az', 'azd', 'terraform', and 'npm'. 4. Sanitization: Relies on the schema validation and processing logic of the official tools invoked.
Audit Metadata