KDAI Node Client

The official client library for connecting nodes to the KDAI distributed AI platform.

Features

• Simple node registration and authentication
• Secure WebSocket-based communication with KDAI hub
• Task management for AI workloads
• Resource utilization monitoring
• Simple CLI interface
• Robust HTTP/HTTPS handling with improved connectivity

Installation

pip install kdai-node-client

Quick Start

Register a New Node

To register a new node with your KDAI server:

# Register a new node
kdai-node register --server-url https://your-kdai-server.com --name "My AI Node"

This will generate an authentication token for your node and save the configuration.

Start the Node

Once registered, you can start the node:

# Start the node
kdai-node start

The node will connect to the KDAI server, report its system information, and start accepting tasks.

Advanced Usage

Monitor Node Status

Check the status of your node:

kdai-node status

Stop Node

Stop a running node:

kdai-node stop

Programmatic Usage

You can also use the KDAI Node Client as a library in your Python code:

from kdai_node_client import KDAINode

# Create a node instance
node = KDAINode(server_url="https://your-kdai-server.com", node_name="My AI Node")

# Register the node
node.register()

# Start the node
node.start()

# To stop the node
node.stop()

API Reference

Full API documentation is available at https://docs.kdai.io/node-client/

Requirements

• Python 3.8 or higher
• Working network connection to KDAI server
• WebSocket support

Recent Changes

Version 0.1.1 (2025-03-27)

• Fixed HTTP to HTTPS redirect handling to improve connection reliability
• Added proper error handling for network connectivity issues
• See CHANGELOG.md for the full history of changes

License

MIT License