Can I test my MCP server locally?

Yes, the skill includes guidance on local setup, including using tools like ngrok to provide the HTTPS tunnel required by ChatGPT during the development phase.

Which transport protocol is recommended for production?

Streamable HTTP is the recommended transport protocol for production deployments as it is designed for publicly accessible servers and reliable communication with the OpenAI Apps SDK.

How do I handle UI widgets using this architecture?

The architecture guides you on using the _meta field and resource URIs within tool responses to provide the structured data and HTML templates required for ChatGPT to render widgets.

What is an MCP server in the context of OpenAI Apps?

An MCP (Model Context Protocol) server acts as the backend for ChatGPT apps, exposing specific tools that the model can invoke to perform actions, fetch data, or trigger UI widgets.

Does this skill support both Python and TypeScript?

Yes, it provides detailed setup instructions, minimal code structures, and implementation patterns for both the Python and TypeScript MCP SDKs.

MCP Server Architecture

Name: MCP Server Architecture
Author: mberto10

bymberto10

API開発

Provides architectural guidance and implementation patterns for building Model Context Protocol (MCP) servers with the OpenAI Apps SDK.

概要

This skill serves as a comprehensive guide for developers building the backend infrastructure for ChatGPT apps. It offers standardized patterns for implementing the Model Context Protocol (MCP), covering critical areas such as transport protocols (Streamable HTTP and SSE), multi-language SDK setup in Python and TypeScript, and the structured response formats required for UI widgets. By following these best practices, developers can ensure their MCP servers handle tool calls, authentication, and data flow efficiently while remaining compatible with OpenAI's production requirements.

主な機能

Standardized tool response patterns for model narration and UI widget rendering.
Best practices for secure hosting, environment variables, and error handling.
0 GitHub stars
Comprehensive SDK setup guides for both Python (FastMCP) and TypeScript.
Production-ready templates for Streamable HTTP and SSE transport protocols.
Optimized project structure recommendations for scalable MCP deployments.

ユースケース

Developing a custom backend for ChatGPT apps to expose internal tools and APIs.
Configuring secure, production-grade MCP servers using Streamable HTTP.
Implementing rich UI components in ChatGPT via MCP resources and metadata.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add mberto10/mberto-compound mcp-server-architecture

For use in Claude.ai and ChatGPT

Download Skill

GitHub

概要