An MCP-enabled 3D Worker Monitoring and Control System
Worker17 is a comprehensive system that allows you to:
The system consists of:
Worker17 started as a practical joke, and then it turned into an exploration of SSE MCP servers. It's a silly project, but it's also a fun way to learn about MCP.
# Start the server
cd server
npm install
npm start
# In another terminal, start the webapp
npm install
npm start
For development and demo purposes, you can run the server directly in the browser using WebContainer:
# Development with WebContainer (builds server automatically)
npm run dev:wc
# Build for production with WebContainer
npm run build:wc
# Preview the production build with WebContainer
npm run preview:wc
This mode automatically builds the server code and runs it in a browser-based Node.js environment, eliminating the need for a separate server process.
You can also run the system using Docker:
docker-compose up
Note: If you are running docker in WSL without Docker Desktop, there's currently a bug that prevents the ports from being exposed. The workaround is to run the container with host network mode. However, this poses potential risk, as it bypasses the networking isolation of the container.
You can use the MCP Inspector to connect to the Worker17 MCP server and inspect the state of the workers. Choose SSE transport and provide the worker17 server URL: http://localhost:4000/sse
(or whatever is the URL you are running the server on).
Claude Desktop currently does not support SSE MCP server. To use the Worker17 MCP server, you will need a proxy Stdio MCP server. I've tested this with mcp-proxy. You will need to install it as a Windows app and put it in the path. I used Windows binaries for uv which added it to %USERPROFILE%.local\bin
Then you can add the worker17 MCP server to Claude Desktop configuration:
{
"mcpServers": {
"worker17": {
"command": "mcp-proxy",
"args": ["http://localhost:4000/sse"]
}
}
}
Make sure the server is up and running before you start Claude Desktop.
{
"mcpServers": {
"worker17": {
"env": {},
"args": [
"http://localhost:4000/sse"
],
"command": "mcp-proxy"
}
}
}
Seamless access to top MCP servers powering the future of AI integration.