microsoft-docs
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- PROMPT_INJECTION (SAFE): No evidence of malicious instructions or attempts to override system prompts. The instructions provided are focused on optimizing search queries and tool usage for the intended documentation task.
- DATA_EXFILTRATION (SAFE): No file access or network exfiltration patterns detected. The tool definitions imply fetching data from Microsoft sources, which is consistent with the skill's purpose.
- EXTERNAL_DOWNLOADS (SAFE): No external scripts, packages, or binary downloads are referenced in the skill definition.
- INDIRECT_PROMPT_INJECTION (LOW): While the skill ingests external content from Microsoft Learn, this is a standard documentation retrieval surface.
- Ingestion points: External data enters through
microsoft_docs_searchandmicrosoft_docs_fetchtools. - Boundary markers: Not explicitly defined in this documentation file.
- Capability inventory: Data retrieval from external web sources.
- Sanitization: Not specified in the markdown documentation, though typical for such integrations.
Audit Metadata