K8S-LLM

中文

A command-line tool that allows users to interact with Kubernetes using natural language, powered by Large Language Models.

Features

• Natural language interface for Kubernetes commands
• Context-aware conversations
• Configurable LLM settings
• Command history and context management

Installation

git clone git@github.com:yexia553/k8sllm.git
cd k8sllm
pip install -r requirements.txt && pip install . e

Usage

# Ask a question
k8sllm -q "list all pods in the default namespace"

# Clear conversation context
k8sllm -c

You can also set alias to make usage easier.

alias qk8sllm='k8sllm -q'
alias ck8sllm='k8sllm -c'

After setting the above alias, you can directly use qk8sllm and ck8sllm.

# Ask a question
qk8sllm "List all pods in the default namespace"

# Clear conversation context
ck8sllm

Configuration

Create a .k8sllm/config.yaml file in your home directory with the following structure:

llm:
  base_url: "https://api.deepseek.com/v1" # Change this to your LLM service URL
  api_key: "your-api-key"
  model: "deepseek-chat" # Change this to your model name

Any large model API compatible with the OpenAI SDK can be used. Personally, I recommend DeepSeek as it offers good performance, is cost-effective, and has a simple configuration.

Development

This project uses Python 3.8+ and requires the following main dependencies:

• click: For CLI interface
• pyyaml: For configuration management
• openai: For LLM integration

License

MIT License