azure-anyscale-on-azure
Pass
Audited by Gen Agent Trust Hub on Oct 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to fetch and process external documentation content from Microsoft Learn using tools like
mcp_microsoftdocs:microsoft_docs_fetchorWebFetch. This creates a vulnerability surface where the agent could ingest instructions embedded in the external content. - Ingestion points: Documentation fetched from URLs under
learn.microsoft.com(referenced in SKILL.md). - Boundary markers: Absent; there are no instructions to the agent to disregard instructions found within the fetched content.
- Capability inventory: The skill utilizes network-enabled tools to fetch markdown content which the agent then parses and acts upon.
- Sanitization: No sanitization or validation of the fetched content is described.
- [EXTERNAL_DOWNLOADS]: The skill references and retrieves configuration guidelines and technical documentation from Microsoft's official documentation service. This is a well-known and established service.
Audit Metadata