Mini Course Generator (MCG) is a learning content platform that turns documents, notes, or a simple prompt into a structured, shareable mini course. With our MCP (Model Context Protocol) connector, you can plan, generate, and publish courses directly from any MCP-compatible AI client — without leaving the conversation.
This guide walks you through what the connector does, how to connect it from popular AI tools, and how to get support if something goes wrong.
What is MCP?
The Model Context Protocol (MCP) is an open standard that lets AI assistants and developer tools securely connect to third-party services. Once an MCP server is connected, the AI can call its tools, read its resources, and surface its UI directly inside the chat.
Mini Course Generator exposes an MCP server so any compatible client — Claude, Cursor, GitHub Copilot, ChatGPT, Windsurf, and others — can generate and publish courses on your behalf.
What you can do with the MCG connector
Once connected, you can ask your AI assistant to:
Generate a course outline from a prompt or from documents you attach
Edit the outline conversationally before publishing
Create a full course from an approved outline, complete with a public, shareable URL
Embed and preview the generated course directly inside supported clients
The connector is designed for educators, trainers, L&D teams, marketers, customer education specialists, and anyone who needs to turn raw material into polished, interactive learning experiences.
Example prompts
"Create a course outline from this onboarding PDF for new sales hires."
"Turn this outline into a published mini course I can share with my team."
"Generate a 5-lesson beginner course on prompt engineering with quizzes."
Requirements
Before you connect, make sure you have:
A Mini Course Generator account — sign up at minicoursegenerator.com
An MCP-compatible AI client (see the list below)
A modern browser to complete the one-time sign-in
MCP server URL:
https://mcp.minicoursegenerator.com/mcp
How to connect
The setup steps differ slightly by client, but the underlying flow is the same: point the client at the MCP URL above, then sign in to your MCG account when prompted.
Claude (Claude.ai web, Claude Desktop)
1. Open the app and go to Settings → Connectors.
2. Click Add custom connector.
3. Paste https://mcp.minicoursegenerator.com/mcp as the server URL.
4. Click Connect. You'll be redirected to the MCG login page — sign in with your MCG account.
5. After login, you'll be redirected back and the connector will be ready to use.
Claude Code (CLI)
Bash command:
claude mcp add --transport http mcg https://mcp.minicoursegenerator.com/mcp
Claude Code will open a browser window for the MCG sign-in the first time you call an MCG tool.
Cursor
1. Open Settings → MCP (or edit `~/.cursor/mcp.json` directly).
2. Add a new server entry:
{
"mcpServers": {
"mini-course-generator": {
"url": "https://mcp.minicoursegenerator.com/mcp"
}
}
}3. Restart Cursor. The first MCG tool call will trigger an OAuth sign-in.
ChatGPT (Developer Mode / Custom Connectors)
1. In ChatGPT, open the connector or app settings (availability depends on your ChatGPT plan).
2. Add a new MCP server with URL https://mcp.minicoursegenerator.com/mcp.
3. Complete the OAuth flow when prompted.
Available tools
mcg_generate_course_outline: Generates a structured course outline from a prompt and optional attached documents. Returns an editable outline.
mcg_create_course_from_outline: Takes an approved outline and produces a complete, publishable course with a public URL and embed-ready widget.
In clients that render MCP UI (such as Claude and ChatGPT), the outline and finished course appear as interactive widgets inline. In text-only clients, the same data is returned as structured JSON your AI assistant can read and summarize.