The Agent Skills Directory

[SAFE]: The analysis did not detect any evidence of malicious code, obfuscation, or persistence mechanisms. The skill operates within expected documentation retrieval parameters.
[EXTERNAL_DOWNLOADS]: Fetches documentation from Microsoft's official learn.microsoft.com domain and references a repository under the MicrosoftDocs GitHub organization.
[PROMPT_INJECTION]: The skill includes instructions labeled 'IMPORTANT' that guide the agent on how to use specific tools and reference sections; these are benign operational directives rather than attempts to bypass safety guardrails.
[INDIRECT_PROMPT_INJECTION]:
Ingestion points: External documentation is fetched using fetch_webpage and mcp_microsoftdocs:microsoft_docs_fetch in SKILL.md.
Boundary markers: The skill does not implement specific boundary markers or instructions to ignore embedded content for ingested data.
Capability inventory: The skill is limited to reading documentation and does not possess capabilities for file writing, subprocess execution, or privilege escalation.
Sanitization: No content sanitization or filtering processes are explicitly defined.

azure-anomaly-detector