deploying-to-azure
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines a pattern for delegating the analysis of container logs and deployment output to a subagent. This presents an indirect prompt injection surface as logs could contain malicious instructions designed to manipulate the subagent's analysis or output.
- Ingestion points: Container logs and deployment output retrieved via
az containerapp logs show(referenced in the Delegation Pattern). - Boundary markers: Absent in the suggested subagent prompt; logs are processed directly without delimiters or explicit isolation.
- Capability inventory: The skill utilizes
az(Azure CLI),azd(Azure Developer CLI), and PowerShell for resource management and infrastructure updates. - Sanitization: No filtering or sanitization of external log content is described before subagent processing.
- [EXTERNAL_DOWNLOADS]: The Dockerfiles (
frontend.Dockerfile,backend.Dockerfile) and deployment hooks utilize standard package managers (npm ci,dotnet restore) to fetch dependencies from public registries. These are standard operations for the skill's stated purpose. - [COMMAND_EXECUTION]: The skill involves the execution of various CLI tools, including
az,azd, and PowerShell, to manage Azure infrastructure, update RBAC assignments, and configure container environments.
Audit Metadata