The Agent Skills Directory

[SAFE]: The skill provides standard documentation for Azure AI services management and implementation.
[SAFE]: Authentication is performed securely via environment variables or Azure Identity SDKs, avoiding hardcoded secrets.
[SAFE]: Network communications are restricted to trusted Azure endpoints for AI and search services.
[SAFE]: Command execution is limited to official Azure CLI tools for infrastructure management.
[SAFE]: RAG pattern implementation uses search results as context. Ingestion points: user query and search results in SKILL.md; boundary markers: newlines; capability inventory: OpenAI chat completion; sanitization: best practice guidance included.

azure-ai