The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill fetches documentation content from the official learn.microsoft.com domain. These references are used to provide the agent with the most up-to-date technical information from the vendor and are considered safe.
[INDIRECT_PROMPT_INJECTION]: The skill has the capability to ingest data from external web pages, which represents a theoretical attack surface for indirect prompt injection.
Ingestion points: Technical documentation fetched from learn.microsoft.com via the microsoft_docs_fetch or fetch_webpage tools.
Boundary markers: The instructions do not specify explicit delimiters or "ignore instructions" prompts for the fetched content.
Capability inventory: The skill is configured to use file reading tools (read_file) and network fetching tools.
Sanitization: No specific sanitization or filtering logic is described for the retrieved documentation content.
Context: The security risk is assessed as safe because the data ingestion is limited to official, trusted Microsoft documentation sources, which are unlikely to contain malicious instructions targeting the agent.

azure-maps