Autonomous Short-Form Video Factory with n8n + GPT

An end-to-end automation pipeline that creates and publishes vertical videos (Reels/Shorts/TikToks) across multiple social media platforms with zero human intervention.

🎯 Project Overview

This project implements a fully automated content creation system that:

• Scrapes trending AI/tech content from TLDR.tech
• Generates engaging video scripts using AI
• Creates vertical videos with voice-overs, music, and captions
• Publishes to Instagram, TikTok, and YouTube Shorts
• Operates completely autonomously

🏗️ Architecture

The system consists of:

• n8n: Workflow automation platform
• crawl4ai: Web scraping service
• Ollama: Local AI model for content generation
• Docker: Containerized deployment
• Colima: Local Docker environment for macOS

📋 Prerequisites

• Docker
• Colima (Only for MacOS)

At least 4GB RAM available for containers

🚀 Quick Start

1. Clone the Repository

git clone git@github.com:migush/ClipClap-Factory.git
cd ClipClap-Factory

2. Start n8n and crawl4ai

MacOS Users: Use Colima

chmod +x start_colima.sh
./start_colima.sh

This script:

• Stops any running Colima instance
• Starts a new Colima environment with 2 CPUs and 4GB RAM
• Mounts your project directories for file sharing

3. Windows/Linux

docker-compose up -d

This will start:

• n8n on http://localhost:5678
• crawl4ai on http://localhost:11235

4. Access n8n

1. Open your browser and navigate to http://localhost:5678
2. Complete the initial n8n setup
3. Import the workflow from ClipClat_Factory.json

5. Configure API Keys (Not Implemented Yet)

📁 Project Structure

ClipClap-Factory/
├── docker-compose.yml           # Container orchestration
├── start_colima.sh              # Colima startup script
├── ClipClat_Factory.json                # Main workflow
├── files/                       # Generated content storage (mouted)
└── .n8n/                        # n8n configuration (mounted)

🔧 Current Workflow

The existing workflow (ClipClat_Factory.json) performs:

1. Content Discovery: Scrapes AI news from TLDR.tech
2. AI Processing: Uses Ollama to generate video scripts
3. Content Storage: Saves generated scripts to files/ directory

🎬 Usage

Running the Current Workflow

1. Open n8n at http://localhost:5678
2. Create a new workflow and import ClipClat_Factory.json
3. Navigate to the imported workflow
4. Make necessary adjustments (e.g., API keys)
5. Execute Workflow
6. Check the files/ directory for generated content

🚧 Development Roadmap

Phase 1: Foundation & Content Discovery ✅

• Infrastructure Setup: Docker containers for n8n and crawl4ai
• Content Scraping: Automated scraping from TLDR.tech AI section
• AI Integration: Ollama local AI model for content processing
• Workflow Automation: n8n workflow for content discovery and processing
• File Management: Automated file generation and storage system

Phase 2: Content Generation & Processing ✅

• AI-Powered Script Generation: Automated script creation from scraped content
• Image Generation: AI-generated images for video content (OpenAI API integration)
• Video Generation: Automated video creation with Kling AI
• Content Processing Pipeline: End-to-end content transformation workflow

Phase 3: Video Enhancement & Production 🔄

• Text-to-Speech Integration: Automated voice-over generation
• Background Music: AI-generated or curated background music
• Caption Overlay System: Automated subtitle generation and placement
• Video Editing: Automated trimming, transitions, and effects
• Quality Optimization: Video compression and format optimization

Phase 4: Multi-Platform Publishing 🎯

• Instagram Integration: Automated posting to Instagram Reels
• TikTok API Integration: Automated posting to TikTok
• YouTube Shorts: Automated posting to YouTube Shorts
• Cross-Platform Optimization: Platform-specific video formatting
• Scheduling System: Intelligent posting time optimization

Phase 5: Advanced Automation & Intelligence 🚀

• Trending Topic Discovery: AI-powered trend analysis
• Content Performance Analytics: Automated performance tracking
• Viral Prediction: AI algorithms for content virality prediction
• A/B Testing: Automated content variation testing
• Audience Engagement: Automated response and interaction management

Phase 6: Enterprise Features & Scaling 🌟

• Multi-Account Management: Support for multiple social media accounts
• Content Calendar: Advanced scheduling and content planning
• Team Collaboration: Multi-user workflow management
• Advanced Analytics: Comprehensive reporting and insights
• API Integration: Third-party service integrations

Current Status: Phase 2 Complete ✅

The system currently successfully:

• Scrapes trending AI content from TLDR.tech
• Generates AI-powered scripts using Ollama
• Creates AI-generated images using OpenAI
• Produces videos using Kling AI
• Stores all generated content automatically

Next Milestone: Phase 3 - Adding audio, music, and caption features to complete the video production pipeline.

🐛 Troubleshooting

Common Issues

Colima won't start

# Check if Docker is running
docker ps

# Restart Colima
colima stop
colima start --cpu 2 --memory 4

n8n not accessible

# Check if containers are running
docker-compose ps

# Restart services
docker-compose down
docker-compose up -d

Logs

View container logs:

# n8n logs
docker-compose logs n8n

# crawl4ai logs
docker-compose logs crawl4ai

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

🙏 Acknowledgments

• n8n for workflow automation
• crawl4ai for web scraping
• Ollama for local AI processing
• Docker for containerization
• Colima for local Docker on macOS

📞 Support

For questions or issues:

1. Review n8n documentation
2. Review crawl4ai documentation
3. Review Ollama documentation
4. Review Docker documentation
5. Review Colima documentation
6. Open an issue on GitHub

---

Happy automating! 🚀