Enterprise-grade automation architectures for AI Engineers, Agencies, and Freelancers.
The Problem: Most automation tutorials use "Black Box" tools that don't scale. Our Solution: This repository provides "Glass Box" workflows using n8n. We break complex tasks (like video generation) into an assembly line of specialized AI agents. You own the code, the API keys, and the data.
Choose your architecture: Pro (Cloud APIs for speed) or Local (Python/Ollama for privacy and zero cost).
| ID | Agent/Module Name | Business Use Case (ROI) | Architecture |
|---|---|---|---|
| 01 | YouTube Documentary Generator | Content Automation: Generates 10-minute, scene-by-scene documentary videos from a single prompt. | ✅ Cloud (API) ✅ Local (Python) |
| 02 | Omni-Channel Social Agent | Marketing: Repurposes 1 Blog/Video URL into formatted LinkedIn, X, and Instagram posts automatically. | ✅ Cloud (Buffer) ✅ Local (Pillow) |
| 03 | B2B Lead Scraper & Enricher | Sales: Crawls Team Pages, permutes emails, and verifies leads for outreach campaigns. | ✅ Cloud (Hunter) ✅ Local (Python DNS) |
| 04 | Self-Healing Error Handler | DevOps: Diagnoses workflow crashes and auto-fixes or alerts via Slack/Discord. | ✅ Cloud (Slack) ✅ Local (Discord) |
Why this repository is different:
We reject "wrapper" tools. We focus on Component-Based Architecture.
- Granular Control: You own every step of the process.
- Scalability: Swap out models (e.g., GPT-4o to Claude 3.5) without breaking the system.
- Cost Efficiency: Toggle between high-performance APIs or local compute power based on client budget.
To use these workflows, you need an instance of n8n.
- Cloud: n8n.io (Easiest for Pro Workflows).
- Self-Hosted: Docker or Pinokio (Required for Local/Python Workflows).
- Navigate to the specific Module Folder (linked in the table above).
- Read the module's
README.mdfor specific API requirements (OpenAI, Anthropic, etc.). - Download the
.jsonworkflow file. - Import to n8n: Open n8n Editor → Top Right Menu → Import from File.
- Orchestration: n8n
- LLMs: OpenAI (GPT-4o), Google Gemini 1.5 Flash, Ollama (Llama 3)
- Vision: Replicate (Flux.1), Hugging Face (SDXL)
- Audio: ElevenLabs, OpenAI TTS
- Compute: Python 3.10 (MoviePy, Pandas)
Maintained by Solih Ahmad Sohan
This is an open-source project. If you find a bug or want to add a "Local" alternative to a "Pro" workflow:
- Fork the repo.
- Create a branch:
git checkout -b feature/amazing-feature. - Open a Pull Request.
Looking for a custom automation for your agency? Connect with me on LinkedIn.