Diagnosing and resolving issues in Kubernetes clusters can
be complex and time-consuming, especially when navigating logs, telemetry, and
diagnostics manually. To simplify this experience, Azure Kubernetes Service is
introducing a new agentic AI-powered CLI experience through the “az aks agent” command.
This new capability brings agentic reasoning directly into
the Azure CLI, allowing developers and operators to interact with their
clusters using natural language. By describing an issue or query, users can
trigger an intelligent agentic loop that gathers signals from multiple sources,
analyzes diagnostics, identifies health issues and root causes, and provides
actionable recommendations. The CLI agent also integrates with the AKS MCP
server and can seamlessly create and manage an Azure OpenAI instance, though it
supports any LLM provider.
Powered by HolmesGPT - an open-source project co-maintained by the AKS team and accepted as a CNCF sandbox project- this feature is now available in public preview.
Learn more.