The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill defines a mechanism for ingesting external documentation, which constitutes a potential attack surface for indirect prompt injection.
Ingestion points: Markdown content is retrieved from learn.microsoft.com using the mcp_microsoftdocs:microsoft_docs_fetch or fetch_webpage tools.
Boundary markers: No specific delimiters or instructions for isolating external content from agent instructions are provided in the skill.
Capability inventory: The skill utilizes tools for fetching remote content and reading local file segments.
Sanitization: The skill does not specify any sanitization or validation of the fetched documentation content.
Risk Assessment: As the content is sourced exclusively from official Microsoft documentation repositories, the risk is negligible.
[EXTERNAL_DOWNLOADS]: The skill provides numerous references to official Microsoft documentation (learn.microsoft.com) and a link to a Microsoft GitHub repository for tool installation. These resources are well-known, trusted, and consistent with the skill's stated purpose.

azure-backup