ClientAgentJS

Documentation

Welcome to the ClientAgentJS documentation. This library is designed to add AI agent capabilities to your web applications with a Zero-Backend approach.

Core Documentation

Integration Features

Practical Examples

Explore these examples to learn how to integrate ClientAgentJS into different scenarios:

Chat Explorer

The flagship example. A professional chat interface that includes session management, streaming cancellation, Markdown rendering, and real-time event logging.

Form Assistance

AI-powered forms. Demonstrates how to add AI help to standard text fields for content improvement and generation based on form context.

MCP Tools

Model Context Protocol. A technical example on connecting the agent to external MCP servers to access databases or local files.


How to run the examples

To test any example locally:

  1. Build the library: npm run build
  2. Serve the project root with a static server (e.g., npx serve or python3 -m http.server)
  3. Open the example path in your browser.

See license: ClientAgentJS