Building Custom AI Agents with JavaScript, React, and ReExt

At JS Days 2026, Marc Gusmano, Sales Engineer at Sencha, will show developers how to build AI agents using JavaScript, React, and ReExt.

Session Overview

Title: Crafting Custom AI Agents with JavaScript, React and ReExt
Speaker: Marc Gusmano
Company: Sencha

This technical session explores how developers can create custom AI agents as Visual Studio Code extensions that integrate naturally into existing development workflows. Rather than relying solely on standard LLM API calls, attendees will learn how modern protocols enable AI agents to become active participants in the software development process.

Moving Beyond Traditional AI Assistants

Most AI development tools focus on generating code snippets or answering programming questions. While valuable, these capabilities often lack awareness of the broader project context and have limited interaction with development environments.

This session examines how developers can build AI agents that:

• Understand project architecture and development context
• Operate directly within Visual Studio Code
• Interact with external tools and services
• Access project-specific knowledge and resources
• Support more intelligent and automated development workflows

The result is an AI experience that feels like a native part of the development environment rather than an external assistant.

Understanding Modern AI Integration Protocols

A key focus of the presentation is the emerging standards that make advanced AI integrations possible.

Agent Client Protocol (ACP)

Attendees will learn how the Agent Client Protocol (ACP) provides a consistent framework for building sophisticated conversational experiences that work with compliant large language models. By separating the client experience from the underlying model, developers gain greater flexibility while reducing vendor-specific implementation complexity.

Model Context Protocol (MCP)

The session also explores the Model Context Protocol (MCP) and how it extends AI capabilities beyond language generation.

Using MCP, developers can connect AI agents to:

• External data sources
• Secure enterprise tools
• Local development environments
• Project resources and documentation
• Custom workflows and automation

These integrations allow AI agents to work with real project context, enabling more accurate responses and practical development assistance.

Building with JavaScript, React, and ReExt

Rather than introducing unfamiliar technologies, the session demonstrates how developers can leverage established JavaScript skills to create advanced AI-powered tooling.

Using JavaScript alongside React and ReExt, attendees will see how modern front-end technologies can be used to develop responsive Visual Studio Code extensions with rich user interfaces while integrating seamlessly with AI services and development workflows.

The presentation combines interface development with protocol integration to create a practical, end-to-end example of a custom AI agent.

What You’ll Learn

During this session, attendees will gain insights into:

• Designing AI-powered Visual Studio Code extensions
• Building context-aware developer experiences
• Implementing Agent Client Protocol (ACP)
• Integrating Model Context Protocol (MCP)
• Connecting AI agents to external tools and local resources
• Creating responsive interfaces using JavaScript, React, and ReExt
• Building development tools that extend beyond traditional AI chat experiences

Who Should Attend

This session is ideal for:

• JavaScript developers
• React developers
• Visual Studio Code extension developers
• Software architects
• Developer tooling engineers
• AI application developers
• Engineering teams exploring AI-assisted development

Why Attend

AI-powered developer tools are evolving quickly, and understanding the technologies behind them is becoming increasingly valuable. Rather than focusing on theory, this session demonstrates practical techniques for building intelligent development experiences that integrate directly with existing workflows.

Whether you’re developing internal engineering tools, experimenting with AI integrations, or building extensions for Visual Studio Code, you’ll leave with a clearer understanding of the protocols, architecture, and implementation patterns behind modern AI agents.

Join Marc Gusmano at JS Days 2026 for a live technical session exploring how JavaScript, React, and ReExt can be used to build intelligent, context-aware AI agents that work naturally within today’s development environments.