For the complete documentation index, see llms.txt. Markdown versions of all docs pages are available by appending .md to any docs URL.

Concepts, guides, and reference docs for running declarative AI agents on Kubernetes, from install through day-two operations.

Your complete guide to the AI agent platform for Kubernetes

What is kagent?

kagent is an innovative AI agent platform designed specifically for Kubernetes environments. It empowers developers and operations teams to create intelligent, autonomous agents that can monitor, manage, and automate complex Kubernetes workloads using the power of large language models (LLMs).

kagent was created at Solo.io in 2025 and is a Cloud Native Computing Foundation sandbox project.

Key Features

• AI-Powered Automation - Create intelligent agents that understand natural language and can perform complex Kubernetes operations
• Multi-Provider Support - Works with OpenAI, Anthropic, Google Vertex AI, Azure OpenAI, Ollama, and custom models
• Tool Integration - Supports Model Context Protocol (MCP) tools, built-in Kubernetes tools, and custom HTTP tools
• Agent-to-Agent Communication - Enable sophisticated workflows through A2A (Agent-to-Agent) interactions
• Comprehensive Observability - Built-in tracing and monitoring to understand agent behavior and performance
• Cloud Native - Designed from the ground up to run natively in Kubernetes environments

Why Choose kagent?

Whether you’re looking to automate routine operations, implement intelligent monitoring, or create sophisticated multi-agent workflows, kagent provides the tools and framework to bring AI to your Kubernetes infrastructure. Start with simple automation and scale to complex, intelligent systems that can reason about your cluster’s state and make informed decisions.

Explore the Documentation