N8N Personal Assistant

A Flask-based API service that integrates with OpenAI's GPT models to provide AI assistance through n8n workflows. This project leverages the gpt-4o-search-preview model, which enables web search capabilities.

Features

• REST API with a /ask endpoint for AI queries
• Integration with OpenAI's GPT models with web search capabilities
• Docker containerization for easy deployment
• Cloudflare Tunnel support for secure public access
• n8n integration for workflow automation

Setup and Deployment

Prerequisites

• Docker and Docker Compose
• OpenAI API key
• n8n instance (for workflow integration)
• Windows or Linux environment

Basic Setup

1. Clone this repository

2. Configure environment variables Create a .env file with the following:

   OPENAI_API_KEY=your_api_key_here
   RUN_API=true
   

3. Install dependencies (for local development only)

   pip install -r requirements.txt
   

Deployment Options

Option 1: Run directly with Python (Development)

python main.py

Option 2: Deploy with Docker (Recommended)

1. Build the Docker image

   Windows:

   docker build -t openai-websearch-api .
   

   Linux:

   sudo docker build -t openai-websearch-api .
   

2. Start the container

   The docker-compose.yml uses host network mode for better compatibility with n8n:

   version: '3'
   
   services:
     openai-websearch-api:
       build: .
       network_mode: "host"
       env_file: 
         - .env
       restart: unless-stopped

   Start the container:

   sudo docker-compose up -d
   

3. Verify deployment

   sudo docker ps
   

   You should see your container running and using the host network.

4. Find your host IP (for n8n configuration):

   ip addr show
   

   Look for your main network interface (usually eth0 or ens4). For example:

   2: ens4: <BROADCAST,MULTICAST,UP,LOWER_UP> ...
       inet 10.138.0.2/32 ...
   

   Note down this IP address (in this example, 10.138.0.2).

n8n Integration

Configure an HTTP Request node in n8n:

1. Set the request details:

   • URL: http://your.host.ip:5000/ask (e.g., http://10.138.0.2:5000/ask)
   • Method: POST
   • Headers: Content-Type: application/json
   • Body:

     {
       "prompt": {{$json.intent}}
     }

2. Process the response in subsequent nodes using {{$json.response}}

Troubleshooting n8n Connection

If you encounter connection issues:

1. Verify the API is running:

   curl -X POST http://localhost:5000/ask -H "Content-Type: application/json" -d "{\"prompt\":\"test\"}"
   

2. Check container logs:

   sudo docker logs $(sudo docker ps -q --filter ancestor=openai-websearch-api)
   

3. Verify network access:

   # Test from the host machine
   curl -X POST http://your.host.ip:5000/ask -H "Content-Type: application/json" -d "{\"prompt\":\"test\"}"
   

4. Common issues and solutions:

   • If n8n can't connect, ensure it's running on the same network or has access to the host network
   • If the host IP changes, update the n8n HTTP Request node URL accordingly
   • For security, consider setting up proper network isolation in production environments

Option 3: Deploy with Docker and Cloudflare Tunnel (Production)

1. Deploy the Docker container (follow Option 2 steps)

2. Set up Cloudflare Tunnel

   Windows:

   • Download cloudflared:

     Invoke-WebRequest -Uri https://github.com/cloudflare/cloudflared/releases/latest/download/cloudflared-windows-amd64.exe -OutFile cloudflared.exe
     

   • Start a tunnel:

     .\cloudflared.exe tunnel --url http://localhost:5000
     

   • Keep the tunnel running in the background:

     Start-Process -NoNewWindow .\cloudflared.exe -ArgumentList "tunnel --url http://localhost:5000"
     

   Linux:

   • Download and install cloudflared:

     curl -L --output cloudflared.deb https://github.com/cloudflare/cloudflared/releases/latest/download/cloudflared-linux-amd64.deb
     sudo dpkg -i cloudflared.deb
     

   • Start a tunnel:

     cloudflared tunnel --url http://localhost:5000
     

   • Keep the tunnel running in the background:

     nohup cloudflared tunnel --url http://localhost:5000 > cloudflared.log 2>&1 &
     

3. Note the assigned URL (e.g., https://something-random-name.trycloudflare.com)

API Usage

Direct API Calls

Send POST requests to the /ask endpoint:

POST /ask
Content-Type: application/json

{
  "prompt": "Your question for the AI assistant"
}

Example with curl:

curl -X POST http://localhost:5000/ask -H "Content-Type: application/json" -d "{\"prompt\":\"What is the current Bitcoin price?\"}"

Example with PowerShell (Windows):

Invoke-WebRequest -Uri "http://localhost:5000/ask" -Method POST -ContentType "application/json" -Body '{"prompt":"What is the current Bitcoin price?"}'

n8n Integration

Configure an HTTP Request node in n8n:

1. Set the request details:

   • URL: http://localhost:5000/ask (local) or https://your-tunnel-url.trycloudflare.com/ask (with Cloudflare)
   • Method: POST
   • Headers: Content-Type: application/json
   • Body:

     {
       "prompt": {{$json.intent}}
     }

2. Process the response in subsequent nodes using {{$json.response}}

VPS Deployment Notes

To deploy on a VPS:

1. Transfer all project files to your VPS:

   Windows to Linux VPS:

   scp -r ./* user@your-vps-ip:/path/to/app/
   

   Linux to Linux VPS:

   rsync -avz --exclude 'venv' --exclude '.git' ./ user@your-vps-ip:/path/to/app/
   

2. Install Docker and Docker Compose on VPS:

   Ubuntu/Debian:

   sudo apt update
   sudo apt install -y docker.io
   
   # Install Docker Compose V2 (plugin method)
   sudo apt install -y docker-compose-plugin
   
   # OR install standalone docker-compose
   sudo curl -L "https://github.com/docker/compose/releases/download/v2.24.6/docker-compose-$(uname -s)-$(uname -m)" -o /usr/local/bin/docker-compose
   sudo chmod +x /usr/local/bin/docker-compose
   
   sudo systemctl enable docker
   sudo systemctl start docker
   

   CentOS/RHEL:

   sudo yum install -y docker
   
   # Install Docker Compose V2 (plugin method)
   sudo yum install -y docker-compose-plugin
   
   # OR install standalone docker-compose
   sudo curl -L "https://github.com/docker/compose/releases/download/v2.24.6/docker-compose-$(uname -s)-$(uname -m)" -o /usr/local/bin/docker-compose
   sudo chmod +x /usr/local/bin/docker-compose
   
   sudo systemctl enable docker
   sudo systemctl start docker
   

3. Follow the Docker deployment steps above (using sudo for all Docker commands)

4. Install cloudflared on the VPS:

   curl -L --output cloudflared.deb https://github.com/cloudflare/cloudflared/releases/latest/download/cloudflared-linux-amd64.deb
   sudo dpkg -i cloudflared.deb
   

5. Run the tunnel command on the VPS:

   nohup cloudflared tunnel --url http://localhost:5000 > cloudflared.log 2>&1 &
   

6. To make the tunnel start automatically on system boot, create a systemd service:

   sudo nano /etc/systemd/system/cloudflared.service
   

   Add the following content:

   [Unit]
   Description=Cloudflare Tunnel
   After=network.target
   
   [Service]
   ExecStart=/usr/local/bin/cloudflared tunnel --url http://localhost:5000
   Restart=always
   User=ubuntu
   
   [Install]
   WantedBy=multi-user.target
   

   Enable and start the service:

   sudo systemctl enable cloudflared
   sudo systemctl start cloudflared
   

System Restart Handling

Automatic Container Restart

The Docker containers are configured to restart automatically on system boot thanks to the restart: unless-stopped directive in docker-compose.yml. This ensures your API service comes back online automatically after a VM restart.

Additional Services

After a system restart, some additional services might need manual restart or verification:

1. Nginx (if used as reverse proxy):

   # Check nginx status
   sudo systemctl status nginx
   
   # Restart nginx if needed
   sudo systemctl restart nginx

2. Verify all services are running:

   # Check Docker containers
   sudo docker ps
   
   # Check nginx
   sudo systemctl status nginx
   
   # Check cloudflared (if using tunnel)
   sudo systemctl status cloudflared

Automating Service Restarts

To ensure nginx restarts automatically after system boot:

# Enable nginx to start on boot
sudo systemctl enable nginx

# Verify it's enabled
sudo systemctl is-enabled nginx

You can also create a simple restart script:

# Create a restart script
sudo nano /usr/local/bin/restart-services.sh

Add this content:

#!/bin/bash
sudo systemctl restart nginx
sudo docker-compose -f /path/to/your/docker-compose.yml up -d
sudo systemctl restart cloudflared  # if using Cloudflare tunnel

Make it executable:

sudo chmod +x /usr/local/bin/restart-services.sh

You can run this script manually after a system restart if needed:

sudo /usr/local/bin/restart-services.sh

Important Notes

• API Security: Consider adding authentication for production deployments
• Cloudflare URL Changes: The free quick tunnel URL changes each time you restart the tunnel
• Rate Limiting: Be mindful of OpenAI API rate limits and costs
• Error Handling: The API includes basic error handling for OpenAI API issues
• Linux Permissions: If running on Linux, ensure proper permissions for Docker and cloudflared
• Docker without sudo: To run Docker without sudo on Linux, add your user to the docker group:

  sudo usermod -aG docker $USER
  # Log out and log back in for changes to take effect
  

• Network Mode: Using host network mode provides better compatibility with n8n but may not be suitable for all production environments
• Security: When using host network mode, your API is accessible on your host's network interface. Consider implementing additional security measures
• IP Address: The host IP may change after system restart on some cloud platforms. Update your n8n configuration accordingly