The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill facilitates fetching external documentation from learn.microsoft.com using tools like microsoft_docs_fetch or fetch_webpage. This establishes a surface where the agent ingests untrusted data that could contain instructions.
Ingestion points: Remote Markdown content retrieved from URLs under https://learn.microsoft.com/en-us/azure/osconfig/.
Boundary markers: The instructions do not specify any delimiters or safety markers to differentiate between the skill's instructions and the fetched documentation.
Capability inventory: The skill does not include any capabilities for subprocess execution, file writes, or network exfiltration to non-whitelisted domains.
Sanitization: The instructions do not provide methods for escaping or filtering the fetched content before it is processed by the agent.
[EXTERNAL_DOWNLOADS]: The skill fetches documentation content from Microsoft's official Learn domain and references a repository under the MicrosoftDocs GitHub organization.
[SAFE]: All documentation sources and referenced tools belong to the official Microsoft namespace (microsoft.com) or the author's official GitHub organization (github.com/MicrosoftDocs). No obfuscation, persistence, or credential harvesting patterns were detected.

azure-osconfig