azure-copilot

Pass

Audited by Gen Agent Trust Hub on Mar 10, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides structured metadata and a category index for Azure Copilot documentation. It does not contain executable code or hidden instructions.
  • [EXTERNAL_DOWNLOADS]: The skill references documentation hosted on official domains (learn.microsoft.com) and recommends installing a tool from the author's own GitHub repository (MicrosoftDocs/mcp). These are trusted vendor resources.
  • [PROMPT_INJECTION]: Analysis of instructions to the agent revealed only benign guidance regarding file reading and version checking. No bypass or override patterns were detected.
  • [DATA_EXFILTRATION]: No access to sensitive system files or credentials was found. Network activity is limited to fetching public documentation.
  • [COMMAND_EXECUTION]: The skill does not perform any local command execution or shell operations.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 10, 2026, 01:41 AM