appfactory-v2-mcp

nelc/appfactory-v2-mcp

3.2

If you are the rightful owner of appfactory-v2-mcp and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to dayong@mcphub.com.

The NELC App Factory MCP Server enables the generation of production-ready Cloud Run applications directly from the Cursor IDE using natural language.

Tools
1
Resources
0
Prompts
0

NELC App Factory MCP Server

Generate production-ready Cloud Run apps directly from Cursor IDE.

This MCP (Model Context Protocol) server allows you to generate complete, deployable applications using natural language in Cursor.

Features

  • 🚀 One-command deployment to GCP Cloud Run
  • 🗄️ PostgreSQL database (optional, auto-configured)
  • 📦 Cloud Storage (optional, auto-configured)
  • 🔒 HTTPS with SSL (automatic with custom domain)
  • ⚛️ Modern tech stack (Node.js + React)

Quick Start

One-Time Setup (Automatic)

Run this single command:

curl -fsSL https://raw.githubusercontent.com/nelc/appfactory-v2-mcp/main/setup.sh | bash

This automatically:

  • ✅ Checks Node.js and GitHub CLI
  • ✅ Installs MCP tool
  • ✅ Configures Cursor
  • ✅ Authenticates GitHub

Then restart Cursor.

Manual Installation (Alternative)

If you prefer manual setup:

# 1. Install MCP tool
npm install -g git+https://github.com/nelc/appfactory-v2-mcp.git

# 2. Configure Cursor
# Add to ~/.config/cursor/mcp.json (Mac/Linux) or %APPDATA%\Cursor\mcp.json (Windows):
{
  "mcpServers": {
    "app-factory": {
      "command": "app-factory-mcp"
    }
  }
}

# 3. Restart Cursor

Usage

Step 1: Generate Code in Cursor

Start a new chat and say:

Generate my Customer Tracker app using the App Factory tool.

Features:
- Add/edit/delete customers
- Search customers by name
- Export to CSV
- Customer notes

Needs database: yes
Needs file storage: no

Wait for all files to be generated.

Step 2: Deploy

In Cursor terminal:

cd ~/customer-tracker  # or wherever code was generated
./deploy.sh

That's it! The script handles:

  • ✅ Git initialization
  • ✅ GitHub repo creation
  • ✅ Code push
  • ✅ Deployment trigger

Step 3: Add DNS

After ~5 minutes, the workflow outputs:

📋 Add DNS: customer-tracker.futurex.sa → 34.xxx.xxx.xxx

Add this A record in Cloudflare, then visit: https://customer-tracker.futurex.sa

Examples

Simple Todo App

Generate my Todo List app using App Factory tool.

Features:
- Add/delete todos
- Mark complete
- Filter by status

Needs database: yes

File Upload App

Generate my Document Manager app using App Factory tool.

Features:
- Upload documents
- Organize in folders
- Search files
- Download/delete

Needs database: yes
Needs file storage: yes

API Service

Generate my Inventory API using App Factory tool.

Features:
- CRUD operations for inventory
- Stock level tracking
- Low stock alerts
- REST API endpoints

Needs database: yes

After Deployment

The workflow will output DNS configuration:

📋 DNS CONFIGURATION REQUIRED:

Type: A
Name: customer-tracker.futurex.sa
Value: 34.xxx.xxx.xxx
TTL: Auto

Add this record in your DNS provider (Cloudflare), and your app will be live at:

https://customer-tracker.futurex.sa

Troubleshooting

"MCP server not responding"

  • Verify installation: which app-factory-mcp
  • Check Node.js version: node --version (requires 18+)
  • Restart Cursor

"Failed to generate prompt"

  • Check internet connection
  • Verify portal is accessible: 
    curl https://appfactory-v2-portal-621571041797.me-central2.run.app/api/health

"Deployment failed"

  • Check GitHub Actions logs
  • Verify GitHub organization membership
  • Contact admin for organization secrets

Technical Details

Generated Tech Stack

  • Frontend: React 18 + Vite
  • Backend: Node.js 20 + Express
  • Database: PostgreSQL (Cloud SQL)
  • Storage: Google Cloud Storage
  • Deployment: Cloud Run (serverless)

Architecture

Each app gets:

  • 2 Cloud Run services (frontend + backend)
  • Load Balancer with unique static IP
  • Database on shared Cloud SQL instance
  • HTTPS with wildcard SSL certificate

Security & Privacy

What this MCP server does:

  1. Takes your app requirements (name, features)
  2. Sends them to the public portal API
  3. Returns a generated prompt for code generation

Data handling:

  • ✅ No secrets or credentials stored or transmitted
  • ✅ No code sent to external services
  • ✅ No proprietary data leaves your machine
  • ✅ Only app requirements (name, features) sent to portal
  • ✅ Portal is public endpoint (same as web form)

What's NOT in this repo:

  • ❌ GCP credentials
  • ❌ Infrastructure configuration
  • ❌ Deployment secrets
  • ❌ Any sensitive organizational data

Safe to use and safe to be public. ✅

Repository Structure

  • This repo (public): MCP server for business users
  • Infrastructure repo (private): Portal, deployment scripts (admin only)

Support

License

MIT