Create an Online Course With Claude or ChatGPT Using MCP

What MCP lets your assistant do

MCP (Model Context Protocol) is a standard that lets AI assistants use tools in other apps. learnfrom.co's MCP server gives Claude and ChatGPT tools for your courses, so you can describe what you want in plain language and the assistant makes the changes.

• Courses: create, update, publish, unpublish, archive and duplicate courses, and change their settings.
• Lessons: create modules and lessons, edit them and reorder them.
• Videos and files: upload a video, import one from a URL, or import a file from a URL.
• Students: view students and their progress, and invite or remove them.

Before you start

• A learnfrom.co account. You can start with a draft course.
• A Claude plan with custom connectors. On a Team or Enterprise plan, an admin may need to enable them first.
• A ChatGPT plan with Developer mode. On Business, Enterprise or Edu plans, an admin may need to allow custom connectors in the workspace settings.

Connect Claude

1. Go to claude.ai, open your avatar menu, then Settings and the Connectors tab.
2. Click Add custom connector, name it learnfrom and paste the server address https://www.learnfrom.co/mcp.
3. Click Add. Claude opens the learnfrom sign-in page. Sign in, review the permissions and choose Allow access.

Connect ChatGPT

1. Go to Settings, then Apps, and turn on Developer mode under advanced settings.
2. Choose Create, name the app learnfrom, and paste the server address https://www.learnfrom.co/mcp.
3. Set authentication to OAuth, save, then sign in on the learnfrom page and choose Allow access.

Build the course in a conversation

Start by asking the assistant to check your account and existing courses, so it can see what's already there. Then give it the outline. A typical order of work is:

1. Create the course with its title, description and price.
2. Add modules, then the lessons inside each module.
3. Upload the lesson videos from a file or a URL, then check the upload status.
4. Review the course, and publish it once you're happy with it.

Ask for a draft first and review it in learnfrom before you publish.

What stays under your control

• Permissions: on the sign-in page you can narrow the access. Scopes cover reading courses, changing courses, reading students and inviting or removing students.
• Confirmations: deleting a course, and inviting a large group of students at once, show a preview first. Nothing changes until you confirm.
• Student invitations: these send email and may incur per-student charges, so they need the student-write permission.
• Revoking access: disconnecting the app or revoking the token ends access immediately.
• Audit log: the assistant can read the audit log of changes to your account.

For the full list of tools, parameters and limits, read the MCP developer documentation. To learn more about the connector, see the AI assistants and MCP page.