azure-communication-services
Pass
Audited by Gen Agent Trust Hub on Mar 10, 2026
Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection due to its functionality of fetching external content.
- Ingestion points: External documentation fetched via mcp_microsoftdocs:microsoft_docs_fetch and fetch_webpage.
- Boundary markers: Absent.
- Capability inventory: Network access for data retrieval.
- Sanitization: Absent.
- [EXTERNAL_DOWNLOADS]: The skill fetches content from the well-known learn.microsoft.com service and recommends tools from the author's official GitHub repository (MicrosoftDocs/mcp).
Audit Metadata