The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill fetches documentation from Microsoft's official domain (learn.microsoft.com) using tools like mcp_microsoftdocs or fetch_webpage. This behavior is consistent with the skill's purpose as a documentation assistant.
[PROMPT_INJECTION]: The skill ingests untrusted data from remote URLs, which represents a potential surface for indirect prompt injection.
Ingestion points: Documentation content retrieved from learn.microsoft.com via microsoft_docs_fetch and fetch_webpage.
Boundary markers: No explicit delimiters or warnings to ignore embedded instructions are provided in the instructions.
Capability inventory: The skill is primarily informational and lacks dangerous local capabilities such as file system writing or arbitrary command execution.
Sanitization: No sanitization of the remote content is specified.
[REMOTE_CODE_EXECUTION]: The skill suggests installing the mcp_microsoftdocs tool from the author's official GitHub repository (github.com/MicrosoftDocs/mcp). This is a legitimate tool recommendation for enhancing the skill's capabilities.

azure-deployment-environments