azure-aigateway
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill makes extensive use of the Azure CLI (
az) to perform administrative tasks, including resource group creation, infrastructure deployment via Bicep, and Identity and Access Management (IAM) role assignments (az role assignment create). - [EXTERNAL_DOWNLOADS]: Fetches official Azure OpenAI API specifications from Microsoft's public GitHub repository and references configuration labs from the Azure-Samples organization to facilitate gateway setup.
- [INDIRECT_PROMPT_INJECTION]: Ingests untrusted data through the import of OpenAPI/Swagger specifications from either local files or remote URLs. These specifications are used to define API structures and can influence the generated policy logic, representing a standard attack surface for third-party content.
- [DYNAMIC_EXECUTION]: Generates Azure API Management (APIM) policy XML fragments containing embedded C# expressions at runtime. These expressions are used for complex logic such as mapping API operations to Model Context Protocol (MCP) tools and calculating token metrics.
Audit Metadata