kagent
Kubernetes-native framework for building and managing AI agents.
- Type
- MCP
- Transport
- http
- Open source
- Yes
- GitHub Stars
- ★ 3.1k
- Source
- mcp-github
- Repository
- github.com/kagent-dev/kagent
Overview
kagent is a Kubernetes-native framework designed to simplify the construction, deployment, and management of AI agents. It supports multiple LLM providers and provides tools via an MCP server. The design principles of kagent include scalability, flexibility, observability, and declarative configuration. Users can define agents and tools using YAML files and manage them via CLI and Web UI. Ideal for developers who need to run AI agents in a Kubernetes environment.
Capabilities
- ▪Supports multiple LLM providers
- ▪Provides tools via MCP server
- ▪Kubernetes-native integration
- ▪Declarative configuration
- ▪Observability support
Use cases
Setup
Refer to the official documentation for installation: [Quick Start](https://kagent.dev/docs/kagent/getting-started/quickstart) and [Installation Guide](https://kagent.dev/docs/kagent/introduction/installation).
This information was compiled by AI from public sources and may contain inaccuracies — please refer to the source.
FAQ
Which LLM providers does kagent support?
Supports OpenAI, Azure OpenAI, Anthropic, Google Vertex AI, and more.
How can I add custom tools in kagent?
Custom tools can be added by creating a ToolServer custom resource.
Related skills
gemini-cli
Gemini CLI is an open-source AI agent that brings Gemini's powerful capabilities directly into the terminal.
private-gpt
Provides a complete API layer for local models, supporting RAG, skills, tools, and more.
planning-with-files
File-based persistent planning skill, ideal for AI coding agents and long-running tasks.
scientific-agent-skills
Transform any AI agent into a scientific assistant with 147 ready-to-use research skills.
susi_alexa_skill
A skill that enables question-and-answer interactions between Alexa and Susi AI.
claude-context
Provides code search capabilities for Claude Code, turning the entire codebase into context.