This directory contains practical examples demonstrating how to use ClientAgentJS in different scenarios.

A full-featured chat interface demonstrating sessions, streaming, event logs, and configuration portability.

Key features:

• Multi-turn chat sessions with conversation memory
• Real-time streaming responses with cancellation support
• Markdown rendering of agent responses
• Profile management (OpenAI, Anthropic, Google, Ollama)
• MCP server integration
• Configuration export/import

AI-powered form assistance demonstrating how to add AI help to standard web forms.

Key features:

• Attaching AI actions to form fields
• Generating and improving text content
• Using ask() and stream() as field-level helpers
• Injecting form context into AI requests

Model Context Protocol integration example showing how to connect the agent to external tools.

Key features:

• Configuring and saving MCP servers
• Listing and calling remote tools
• Enabling MCP servers in profiles
• Running AI requests with tool access

For detailed API documentation and architecture information, see:

• Documentation Index
• API Reference
• Architecture Overview

See license: ClientAgentJS