azure-ai-agents-persistent-dotnet
Pass
Audited by Gen Agent Trust Hub on Jun 25, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill documents the standard use of the official Microsoft Azure AI Agents SDK. All referenced resources, including NuGet packages and GitHub repositories, belong to official and trusted Microsoft/Azure domains.
- [EXTERNAL_DOWNLOADS]: References standard .NET packages (Azure.AI.Agents.Persistent, Azure.Identity) from the official NuGet registry. These are well-known, trusted libraries for Azure development.
- [DATA_EXFILTRATION]: No suspicious network operations or exfiltration patterns were detected. Communication is targeted at official Azure resource endpoints (e.g., services.ai.azure.com) which are configured via environment variables.
- [CREDENTIALS_UNSAFE]: Uses secure authentication patterns. The examples correctly instruct the use of DefaultAzureCredential and environment variables instead of hardcoding API keys or secrets.
- [COMMAND_EXECUTION]: Mentions the Code Interpreter tool, which is a standard feature of the Azure AI Agents service allowing for sandboxed Python execution within the Azure environment, rather than arbitrary local command execution.
Audit Metadata