Workflow Engine - Free N8N Alternative

A Python-based workflow automation engine that provides N8N-like functionality for free. Perfect for automating tasks, creating AI workflows, and building YouTube automation pipelines.

🚀 Features

✅ Free and Open Source - No licensing fees
✅ Multiple Node Types - HTTP, AI (OpenRouter/OpenAI), Transform, Conditions, Triggers, YouTube, Database, Image, Voiceover
✅ Async Execution - Fast, non-blocking workflow execution
✅ Easy Configuration - Simple JSON-based workflow definitions
✅ API Key Management - Secure credential handling
✅ Extensible - Easy to add custom node types
✅ YouTube Automation - Complete workflows for AI faceless channels

📋 Requirements

Python 3.8+
aiohttp library

🛠️ Installation

1. Clone the Repository

git clone https://github.com/simplifaisoul/workflow-engine.git
cd workflow-engine

2. Install Dependencies

pip install aiohttp

Or install from requirements:

pip install -r requirements.txt

⚡ Quick Start

1. Configure API Keys (Optional for basic tests)

Create workflow_config.json:

{
  "api_keys": {
    "openrouter": "your-openrouter-api-key-here",
    "openai": "your-openai-api-key-here",
    "github": "your-github-token-here"
  }
}

Or set environment variables:

# Windows PowerShell
$env:OPENROUTER_API_KEY="your-key-here"

# Linux/Mac
export OPENROUTER_API_KEY="your-key-here"

2. Run Quick Test (No API Keys Required)

python workflow_engine/quick_test.py

This tests basic workflow execution without needing any API keys!

3. Create Your First Workflow

Create a file my_workflow.json:

{
  "name": "My First Workflow",
  "nodes": [
    {
      "id": "trigger_1",
      "name": "Start",
      "type": "trigger",
      "parameters": {
        "trigger_type": "manual"
      }
    },
    {
      "id": "ai_1",
      "name": "AI Processing",
      "type": "ai",
      "parameters": {
        "provider": "openrouter",
        "model": "openai/gpt-3.5-turbo",
        "prompt": "Say hello in one sentence"
      }
    }
  ],
  "connections": {
    "trigger_1": ["ai_1"]
  }
}

4. Run the Workflow

# Load workflow
python -m workflow_engine.main load --workflow-id my_workflow --file my_workflow.json

# Run workflow
python -m workflow_engine.main run --workflow-id my_workflow

📚 Documentation

Full Setup Guide - Complete setup instructions
YouTube Automation Guide - Automate AI faceless YouTube channels
Testing Guide - How to test the engine
Run Tests - Detailed testing instructions
Workflow Engine README - Detailed API documentation

🎯 Use Cases

1. Basic Workflow Automation

HTTP requests
Data transformation
Conditional logic
Database operations

2. AI-Powered Workflows

Script generation
Content creation
Text processing
Image descriptions

3. YouTube Channel Automation

Find viral videos
Generate 3-act story scripts
Create b-roll images
Generate voiceovers
Render videos

See workflow_engine/examples/youtube_automation/ for complete YouTube automation workflows.

🔑 Required API Keys

For Basic Workflows

None required! Basic workflows work without any API keys.

For AI Features

OpenRouter API Key (Recommended)
- Get from: https://openrouter.ai/keys
- Cost: Pay-per-use (~$0.01-0.05 per request)
- Set as: OPENROUTER_API_KEY

For YouTube Automation

OpenRouter API Key (Required for script generation)
YouTube Data API Key (Optional, for fetching video data)
ElevenLabs API Key (Optional, free TTS available)
Image Generation API (Optional, free options available)

See YOUTUBE_AUTOMATION_SETUP.md for complete setup.

📖 Example Workflows

Simple AI Workflow

{
  "name": "Simple AI Workflow",
  "nodes": [
    {
      "id": "trigger_1",
      "type": "trigger",
      "parameters": {"trigger_type": "manual"}
    },
    {
      "id": "ai_1",
      "type": "ai",
      "parameters": {
        "provider": "openrouter",
        "model": "openai/gpt-3.5-turbo",
        "prompt": "Summarize: {{$json.text}}"
      }
    }
  ],
  "connections": {
    "trigger_1": ["ai_1"]
  }
}

HTTP to AI Workflow

{
  "name": "HTTP to AI",
  "nodes": [
    {
      "id": "trigger_1",
      "type": "trigger",
      "parameters": {"trigger_type": "manual"}
    },
    {
      "id": "http_1",
      "type": "http",
      "parameters": {
        "method": "GET",
        "url": "https://api.github.com/repos/microsoft/vscode"
      }
    },
    {
      "id": "ai_1",
      "type": "ai",
      "parameters": {
        "provider": "openrouter",
        "model": "openai/gpt-3.5-turbo",
        "prompt": "Analyze this data: {{$json}}"
      }
    }
  ],
  "connections": {
    "trigger_1": ["http_1"],
    "http_1": ["ai_1"]
  }
}

More examples in workflow_engine/examples/

🧪 Testing

Quick Test (No API Keys)

python workflow_engine/quick_test.py

Validate Setup

python workflow_engine/validate_setup.py

Full Test (Requires OpenRouter Key)

$env:OPENROUTER_API_KEY="your-key"
python workflow_engine/test_workflow.py

🏗️ Architecture

workflow_engine/
├── __init__.py
├── workflow_engine.py    # Core engine
├── config.py             # Configuration management
├── main.py               # CLI interface
├── nodes/                # Node implementations
│   ├── base_node.py
│   ├── http_node.py
│   ├── ai_node.py
│   ├── transform_node.py
│   ├── trigger_node.py
│   ├── condition_node.py
│   ├── youtube_node.py
│   ├── database_node.py
│   ├── image_node.py
│   └── voiceover_node.py
└── examples/             # Example workflows
    ├── simple_workflow.json
    ├── http_to_ai_workflow.json
    └── youtube_automation/
        ├── find_viral_videos.json
        ├── generate_script.json
        ├── generate_images.json
        ├── generate_voiceover.json
        └── main_orchestrator.json

🔧 Available Node Types

Node Type	Description	API Key Required?
`trigger`	Workflow entry points	❌ No
`http`	Make HTTP requests	❌ No
`ai`	AI API interactions (OpenRouter/OpenAI)	✅ Yes
`transform`	Data manipulation	❌ No
`condition`	Conditional logic	❌ No
`youtube`	YouTube operations	❌ No
`database`	File-based database	❌ No
`image`	Image generation	✅ Optional
`voiceover`	Text-to-speech	✅ Optional

💻 Python API Usage

import asyncio
from workflow_engine.workflow_engine import WorkflowEngine
from workflow_engine.config import WorkflowConfigManager

async def main():
    # Initialize
    config = WorkflowConfigManager()
    engine = WorkflowEngine(config)

    # Load workflow
    engine.load_workflow_from_file("my_workflow", "my_workflow.json")

    # Run workflow
    execution_id = await engine.execute_workflow("my_workflow", {"text": "Hello"})

    # Check status
    await asyncio.sleep(2)
    status = engine.get_execution_status(execution_id)
    print(status)

asyncio.run(main())

🎨 Extending the Engine

Add custom node types:

from workflow_engine.nodes.base_node import BaseNode

class MyCustomNode(BaseNode):
    async def execute(self, input_data):
        # Your custom logic here
        return {"result": "custom"}

engine.register_node_type("custom", MyCustomNode)

📊 Cost Breakdown (YouTube Automation)

Per video cost:

Script Generation: $0.01-0.05 (OpenRouter)
Images: $0.00 (free APIs)
Voiceover: $0.00 (free TTS)
Video Rendering: $0.00 (FFmpeg local)
Total: $0.01-0.05 per video

🤝 Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

📝 License

MIT License - feel free to use this project for any purpose.

🆘 Support

⭐ Features Comparison with N8N

Feature	N8N	This Engine
Cost	Paid/Enterprise	Free
Open Source	Partial	Yes
Self-Hosted	Yes	Yes
Python-based	No	Yes
Extensible	Yes	Yes
API Integration	Yes	Yes
AI Integration	Yes	Yes

🎉 Getting Started Checklist

[ ] Install Python 3.8+
[ ] Install dependencies: pip install aiohttp
[ ] Run quick test: python workflow_engine/quick_test.py
[ ] (Optional) Get OpenRouter API key
[ ] (Optional) Create workflow_config.json
[ ] Load example workflow
[ ] Run your first workflow!

Ready to automate! 🚀 Start with the quick test to verify everything works, then build your own workflows!

Specs

README