CourseCode Framework
Open-source framework to build, preview, and package LMS courses with AI tools.
Tested on Linux
Use it
Linux x86_64
You need first
- Runtimes:
Node.js 22.15 or later
install coursecode globally
npm install -g coursecode
Check that it works
coursecode --helpThe output should contain “Start MCP server for AI agent integration”.
Add it to your agent
claude mcp add coursecode -- coursecode mcpJSON for Claude Desktop, Cursor and other clients
{
"mcpServers": {
"coursecode": {
"command": "coursecode",
"args": [
"mcp"
]
}
}
}Create a project with coursecode create <course-name>, then from inside it start the MCP server with coursecode mcp and connect your AI client to it.
What it does
- Build LMS courses with AI toolsTested
How to use it
- Create a course with
coursecode create <course-name>. - In the course folder, start the MCP server with
coursecode mcpand connect your AI client. - Ask the AI to use the
coursecode_component_catalogtool to list available components.
- Create a course with
- Preview, screenshot, lint and test coursesTested
How to use it
- Create a course with
coursecode create <course-name>. - In the course folder, run
coursecode mcpand connect your AI client. - Ask the AI to call the
coursecode_linttool to check the course.
- Create a course with
- Export courses as SCORM, cmi5 and LTI packagesTested
How to use it
- Create a course with
coursecode create <course-name>. - In the course folder, run
coursecode mcpand connect your AI client. - Ask the AI to call the
coursecode_buildtool to build the package.
- Create a course with
How far pickfit got
Linux x86_64
- Installs
- Connects
- 3 of 3 features work
- Works offline
Good for
- Export courses as SCORM, cmi5, or LTI packages
- Preview courses locally with live reload
- Convert PDFs, Word, and PowerPoint to markdown
- Generate audio narration from text
- Let AI preview and screenshot slides to verify layout
How pickfit tested it
Each run starts from a clean machine and follows the steps above. How pickfit verifies
Linux x86_64
An MCP client called one of its tools
- Replayed from scratch
- Works offline
- Checked on 30 Oct 2026
- Version 0.4.4
An MCP client handshake over stdio listed 16 tools, and a call to coursecode_component_catalog returned a result.
Differences the test machine hit
- README: npx coursecode create my-course
In the test: used global install (npm install -g coursecode) then `coursecode create` so the `coursecode mcp` launch command stays on PATH for offline replay