Implement CLI Agent for AKS with AI-powered troubleshooting capabilities

[Copilot]

This PR implements the CLI Agent for Azure Kubernetes Service (AKS) as described in the AKS blog post, providing AI-powered operations and diagnostics for AKS clusters.

Overview

The AKS CLI Agent is an intelligent command-line assistant that helps users troubleshoot, optimize, and operate their AKS clusters with AI-driven insights and automated diagnostics. It integrates seamlessly with the Azure CLI as az aks agent and follows the security model outlined in the blog post.

Key Features

🤖 AI-Powered Troubleshooting

• Natural language query interface supporting questions like "why is my node not ready?" or "diagnose DNS issues"
• Intelligent query classification for targeted responses (node health, DNS, pod scheduling, cost optimization)
• Support for multiple AI providers: Azure OpenAI, OpenAI, and Anthropic Claude

🔒 Security & Privacy

• Local execution: All diagnostics run on the user's machine - no cluster data leaves the client
• Azure CLI authentication: Inherits existing az login session and RBAC permissions
• Bring your own AI: Users configure their preferred AI provider (no data retained by Microsoft)

🛠 Comprehensive Diagnostics

• Node Health: Detect NotReady states, resource pressure, kubelet issues
• DNS Troubleshooting: Identify CoreDNS problems, resolution failures, network policies
• Pod Scheduling: Analyze pending pods, resource constraints, affinity issues
• Cluster Health: Monitor API server, components, and overall cluster status
• Cost Optimization: Provide recommendations for resource efficiency

Usage Examples

# Basic cluster health check
az aks agent "how is my cluster healthy?"

# Node troubleshooting
az aks agent "why is one of my nodes in NotReady state?"

# DNS diagnostics
az aks agent "why are my pods failing DNS lookups?"

# Pod scheduling issues
az aks agent "why is my pod stuck in Pending state?"

# Cost optimization
az aks agent "how can I optimize the cost of my cluster?"

# Check agent status
az aks agent status

# Interactive configuration
az aks agent configure

Architecture

The implementation follows a modular design built on open-source foundations:

• Core Agent Engine: Python-based with extensible AI provider abstraction
• Azure Integration: Native Azure SDK integration with proper authentication
• Kubernetes Client: Direct cluster API access for real-time diagnostics
• Diagnostics Engine: Comprehensive kubectl-based analysis tools
• Configuration Management: YAML-based config with environment variable support

Installation Options

1. Quick Demo: Run ./az-aks-agent.sh --help for immediate testing
2. Python Package: Install with pip install -e cli-agent/
3. Azure CLI Integration: Shell alias or proper extension packaging
4. Direct Usage: python -m aks_agent.cli

Files Added

• cli-agent/: Complete Python package implementation
• az-aks-agent.sh: Azure CLI integration script with demo mode
• cli-agent/README.md: Comprehensive usage documentation
• cli-agent/INSTALLATION.md: Installation guide for different deployment scenarios
• cli-agent/examples/: Configuration templates and usage examples
• cli-agent/tests/: Unit tests for core functionality

This implementation provides the foundation described in the AKS CLI Agent blog post, enabling users to leverage AI-powered assistance for complex Kubernetes troubleshooting scenarios while maintaining security and privacy through local execution.

Fixes #5232.

---

✨ Let Copilot coding agent set things up for you — coding agent works faster and does higher quality work when set up for your repo.