Skip to main content

​
Overview

Genie Helper is a self-hosted AI operations platform for adult content creators. This guide covers installation on a Linux VPS with Plesk.

​
Prerequisites

​
System Requirements

  • OS: Linux (Ubuntu 20.04+ or Debian 11+ recommended)
  • RAM: Minimum 16GB (32GB recommended)
    • Ollama models: ~4.8GB per loaded model
    • Stagehand browser sessions: ~300MB per active session
    • ~10GB available = ~33 concurrent browser sessions max
  • CPU: Multi-core CPU (GPU optional but recommended for faster LLM inference)
  • Disk: 50GB+ SSD storage
  • Server: IONOS dedicated VPS or equivalent (~$100/mo)

​
Required Software

  • Node.js: v18 or higher
  • PM2: Process manager for Node.js services
  • Nginx: Reverse proxy (via Plesk or standalone)
  • Redis: Required for BullMQ job queue
  • Ollama: Local LLM inference engine
  • ImageMagick: Media watermarking
  • FFmpeg: Video processing and teaser generation
  • Git: For cloning the repository

​
Domain Setup

  • Primary domain: geniehelper.com (or your domain)
  • Optional admin subdomains:
    • cms.geniehelper.com - Directus admin panel
    • agent.geniehelper.com - AnythingLLM admin UI

​
Installation Steps

​
1. Install System Dependencies

# Update package list
sudo apt update && sudo apt upgrade -y

# Install Node.js (v18+)
curl -fsSL https://deb.nodesource.com/setup_18.x | sudo -E bash -
sudo apt install -y nodejs

# Install PM2 globally
sudo npm install -g pm2

# Install Redis
sudo apt install -y redis-server
sudo systemctl enable redis-server
sudo systemctl start redis-server

# Install ImageMagick and FFmpeg
sudo apt install -y imagemagick ffmpeg

# Install Git
sudo apt install -y git

​
2. Install Ollama

# Install Ollama
curl -fsSL https://ollama.ai/install.sh | sh

# Start Ollama service
sudo systemctl enable ollama
sudo systemctl start ollama

# Verify Ollama is running
curl http://localhost:11434/api/version

​
3. Pull Required Ollama Models

# Orchestrator / tool planning
ollama pull dolphin3:8b-llama3.1-q4_K_M

# Uncensored content writer
ollama pull dolphin-mistral:7b

# Primary agent model
ollama pull qwen-2.5:latest

# Fallback classifier
ollama pull phi-3.5:latest

# Lightweight summarizer
ollama pull llama3.2:3b

# Fast taxonomy classifier
ollama pull scout-fast-tag:latest

# Embeddings model
ollama pull bge-m3:latest

​
4. Clone Repository

# Clone to your desired location
cd /var/www/vhosts/geniehelper.com/
git clone <repository-url> agentx
cd agentx

​
5. Install Dependencies

# Install root dependencies
npm install

# Install AnythingLLM server dependencies
cd server && npm install && cd ..

# Install Directus dependencies
cd cms && npm install && cd ..

# Install dashboard dependencies
cd dashboard && npm install && cd ..

# Install media worker dependencies
cd media-worker && npm install && cd ..

​
6. Build Dashboard

cd dashboard
npm run build
cd ..
The built files will be in dashboard/dist/.

​
7. Configure Environment Variables

See the Configuration page for detailed environment variable setup.

​
8. Initialize Database

# AnythingLLM Prisma setup
cd server
npx prisma generate
npx prisma migrate dev --name init
cd ..

# Directus initialization (first run)
# Directus will auto-migrate on first startup

​
9. Start Services with PM2

See the Services page for PM2 configuration and startup commands.

​
10. Configure Nginx

See the Nginx Setup page for reverse proxy configuration.

​
Post-Installation

​
Verify Services

# Check all services are running
pm2 status

# Should show 6 processes running:
# - anything-llm (port 3001)
# - agentx-cms (port 8055)
# - stagehand-server (port 3002)
# - genie-dashboard (port 3100)
# - media-worker (no port)
# - anything-collector (no port)

​
Access Admin Panels

ServiceURLDefault Credentials
Dashboardhttps://geniehelper.com/app/(register new account)
Admin Panelhttps://geniehelper.com/admin[email protected]
Directushttp://localhost:8055/admin[email protected] / password
AnythingLLMhttp://localhost:3001[email protected] / (MY)P@$$w3rd
Security Note: Change all default passwords before public launch.

​
Troubleshooting

​
Service Won’t Start

# Check PM2 logs
pm2 logs <service-name> --lines 50

# Example:
pm2 logs anything-llm --lines 50
pm2 logs media-worker --lines 50

​
Redis Connection Issues

# Check Redis status
sudo systemctl status redis-server

# Test Redis connection
redis-cli ping
# Should return: PONG

​
Ollama Model Issues

# List installed models
ollama list

# Test model generation
ollama run qwen-2.5:latest "Hello"

​
Port Conflicts

Ensure the following ports are available:
  • 3001 (AnythingLLM)
  • 8055 (Directus)
  • 3002 (Stagehand)
  • 3100 (Dashboard - if using serve)
  • 11434 (Ollama)
  • 6379 (Redis)
# Check port usage
sudo netstat -tulpn | grep -E '3001|8055|3002|3100|11434|6379'

​
Next Steps

Build docs developers (and LLMs) love

Get started for freeTalk to us