azure-mgmt-apimanagement-dotnet

Pass

Audited by Gen Agent Trust Hub on Mar 1, 2026

Risk Level: SAFEPROMPT_INJECTIONCREDENTIALS_UNSAFEEXTERNAL_DOWNLOADS
Full Analysis
  • [PROMPT_INJECTION]: The skill documentation in 'references/apis-operations.md' shows how to import API definitions from local files or remote URLs. This introduces an indirect prompt injection surface where a malicious OpenAPI specification could contain instructions designed to manipulate the agent.
  • Ingestion points: External data is ingested through 'File.ReadAllText' or 'ContentFormat.OpenApiJsonLink'.
  • Boundary markers: The agent is not provided with markers to isolate imported content from instructions.
  • Capability inventory: The skill allows the creation and modification of Azure service resources and policies.
  • Sanitization: There is no evidence of content sanitization for the imported schemas.
  • [CREDENTIALS_UNSAFE]: The file 'references/products-subscriptions.md' contains a hardcoded dummy password 'SecurePassword123!' in a sample snippet for creating a user account.
  • [EXTERNAL_DOWNLOADS]: The 'SKILL.md' file recommends installing the 'Azure.ResourceManager.ApiManagement' and 'Azure.Identity' packages from NuGet, which are official and trusted libraries from Microsoft.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 1, 2026, 12:35 AM