microsoft-agent-framework

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: LOWEXTERNAL_DOWNLOADS
Full Analysis
  • EXTERNAL_DOWNLOADS (LOW): The skill documentation includes commands to install several .NET packages from NuGet (Microsoft.Agents.AI, Microsoft.Agents.AI.OpenAI, Mscc.GenerativeAI.Microsoft, Azure.Identity). While these are legitimate packages, users should always verify the exact version and source when adding dependencies to a project.
  • INDIRECT PROMPT INJECTION (LOW): The skill provides patterns for processing external inputs (e.g., agent.RunAsync(user_input) and image processing). This creates a standard attack surface for indirect prompt injection. The documentation correctly mitigates high-severity risks by demonstrating 'Human-in-the-Loop' patterns for sensitive operations like tool calls.
  • CREDENTIALS_UNSAFE (SAFE): The provided code examples use placeholders such as <api-key>, <resource>, and <gemini-api-key> rather than hardcoding sensitive credentials. It also promotes the use of AzureCliCredential for safer authentication in local development.
Audit Metadata
Risk Level
LOW
Analyzed
Feb 17, 2026, 12:23 PM