building-chatgpt-apps

Build ChatGPT apps with interactive widgets and FastMCP-backed tools.

Updated Aug 23, 2026
One-click install
npx skills add https://github.com/panaversity/agentfactory --skill building-chatgpt-apps
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: building-chatgpt-apps
Source: https://github.com/panaversity/agentfactory/tree/main/.claude/skills/building-chatgpt-apps
Command: npx skills add https://github.com/panaversity/agentfactory --skill building-chatgpt-apps

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires mcp, and includes scripts (resource) and references (resource) components.

What problem does it solve?

Guides building ChatGPT Apps with embedded widgets, MCP servers, and response metadata to render rich UI inside conversations.

Core Features & Use Cases

  • Widget integration patterns with window.openai APIs (toolOutput, sendFollowUpMessage)
  • MCP server setup (FastMCP), response metadata, and widget embedding
  • Quick-start templates and debugging guidance
  • Use cases: building interactive UX within ChatGPT, chaining tools, and UI-driven prompts

Quick Start

Set up an MCP server with a simple widget; register the widget in ChatGPT; test with a sample app.

Frequently Asked Questions about building-chatgpt-apps

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I build interactive widgets for ChatGPT apps?

Build ChatGPT app widgets by embedding HTML with the text/html+skybridge MIME type and using window.openai APIs like sendFollowUpMessage and toolOutput. The Skill covers widget architecture, MCP server setup with FastMCP, and returning _meta widget resources in tool responses to render rich UI inside conversations.

What is MCP and how do I set up an MCP server for ChatGPT apps?

MCP (Model Context Protocol) enables ChatGPT apps to access external tools and data. Set up an MCP server using FastMCP, register it with ChatGPT, and configure tool endpoints. The Skill guides server setup, response metadata handling, and integrating MCP-backed tools into widget-enabled applications.

Can I add interactive UI components to ChatGPT conversations?

Yes. Add interactive UI by embedding widgets in ChatGPT conversations using the window.openai API for sendFollowUpMessage, toolOutput, and callTool operations. The Skill provides widget integration patterns, templates, and debugging guidance for building rich, interactive user experiences within ChatGPT.

How do I chain multiple tools and pass data between widgets in ChatGPT?

Chain tools by using sendFollowUpMessage and callTool APIs to orchestrate sequential tool calls within widgets. Pass data through tool responses with _meta widget resources and response metadata. The Skill covers tool-chaining patterns and UI-driven prompts for coordinating complex workflows.

What are the requirements for embedding widgets in ChatGPT apps?

Widget embedding requires the text/html+skybridge MIME type, the window.openai API surface (sendFollowUpMessage, toolOutput, callTool), and _meta widget resources in tool responses. An MCP server backend and FastMCP setup satisfy the foundation for registering widgets and handling response metadata in conversations.