azure-cognitive-search
Pass
Audited by Gen Agent Trust Hub on Mar 10, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill acts as a comprehensive index for Azure AI Search technical documentation. It contains no executable scripts, and all logic is confined to providing the agent with navigation instructions for the provided content.- [EXTERNAL_DOWNLOADS]: The skill fetches documentation from 'learn.microsoft.com' and references an official MicrosoftDocs repository on GitHub for the 'mcp_microsoftdocs' server. These are verified, trusted sources and do not represent a security risk.- [DATA_EXFILTRATION]: Network access is used exclusively for fetching technical documentation from well-known official Microsoft domains. No sensitive data is accessed, hardcoded, or transmitted to unauthorized third parties.- [PROMPT_INJECTION]: Instructional content for the agent is focused on managing the context window (reading specific line ranges) and using appropriate tools for documentation fetching. No patterns suggesting behavior overrides or safety bypasses were detected.
Audit Metadata