MCP (Model Context Protocol)
MCP is an open standard that allows AI models to connect with external tools, resources, and services. Think of it as an API layer specifically designed for LLMs. An MCP server exposes:- Tools: Functions the model can call (e.g.,
search_flights,get_weather) - Resources: Data the model can access (e.g., files, UI components)
Two Runtimes, One Codebase
Skybridge supports two runtime environments for rendering widgets:Apps SDK (ChatGPT)
OpenAI’s proprietary SDK for ChatGPT Apps. Uses
window.openai API. Full feature set including modals and file operations.MCP Apps
Open MCP ext-apps specification. Uses JSON-RPC postMessage bridge. Works across multiple AI clients.
Comparison at a Glance
| Feature | Apps SDK (ChatGPT) | MCP Apps |
|---|---|---|
| Protocol | Proprietary window.openai | Open MCP ext-apps spec |
| Client Support | ChatGPT only | Goose, VSCode, Postman, … |
| Specification | Closed | Open source |
What is a ChatGPT / MCP App?
A ChatGPT or MCP App consists of two components:- MCP Server: Handles your business logic and exposes tools
- UI Widgets: React components that render in the AI’s interface
- Text content: What the model sees and responds with
- Widget content: A visual UI that renders for the user
Read our in-depth blog article for a detailed technical breakdown of how ChatGPT Apps work under the hood.