Builder Series: Building an AI-Native Experience with MCP & ChatGPT Apps

​AI apps aren't just chatbots with a better UI.

​They're a new class of software defined by how agents discover, connect to, and act on services, and building them well requires rethinking interface design, user journeys, and architecture.

​Live session agenda

​In this 1-hour live session, we'll explore what it means to build an MCP App or ChatGPT App from first principles:

• ​What are they?

• ​What makes an agentic experience (AX) feel natural?

• ​What separates a demo from something production-worthy?

​We'll then dive into a live build using Skybridge, Alpic's open-source TypeScript framework that lets you build once and deploy across MCP-compatible hosts (Claude, ChatGPT) without duplicating logic.

​What we'll cover:

• ​Core concepts: what defines AI-native apps and agentic user journeys

• ​Live walkthrough of Skybridge's key tools:

  • ​Project scaffolding with the create command

  • ​Dev server with rendering emulator and MCP inspector

  • ​Built-in skills for hassle-free development

  • ​Tunneling to test directly in ChatGPT and Claude

  • ​Hooks library for managing model interactions

  • ​Widget-to-context synchronization for coherent conversations

​What you'll leave with:

• ​A mental model for designing agentic experiences

• ​Understanding of Skybridge's architecture and how to apply it to your projects

• ​Clear next steps for building your own MCP or ChatGPT App