m365-agents-dotnet

Pass

Audited by Gen Agent Trust Hub on Jun 25, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs the user to install official Microsoft .NET packages via NuGet (e.g., Microsoft.Agents.Hosting.AspNetCore, Microsoft.Agents.Authentication.Msal). These packages are hosted on the standard NuGet registry and originate from a trusted organization.
  • [SAFE]: All external references point to official Microsoft documentation (learn.microsoft.com) and the verified Microsoft GitHub repository (github.com/microsoft/agents).
  • [SAFE]: Configuration templates for appsettings.json correctly use curly-brace placeholders (e.g., {{ClientSecret}}) for sensitive credentials, preventing accidental hardcoding of secrets.
  • [PROMPT_INJECTION]: The skill implements an indirect prompt injection surface in the OnMessageAsync handler, where user-supplied text from turnContext.Activity.Text is echoed back. This is the intended primary purpose of a chatbot framework and does not represent a malicious pattern in this context.
  • Ingestion points: turnContext.Activity.Text in MyAgent class.
  • Boundary markers: None (typical for basic agent samples).
  • Capability inventory: Sending messages back to the user via turnContext.SendActivityAsync.
  • Sanitization: None (standard for boilerplate echo agents).
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 25, 2026, 12:40 PM
Security Audit — agent-trust-hub — m365-agents-dotnet