Workflows
Workflows are the foundation of WRLDWIDE AI. They let you chain AI operations, process content at scale, and automate repetitive tasks — all through a visual interface. The platform supports two workflow modes: Dark Factory for pre-built content pipelines, and the Node System for custom workflows.Workflow Types
Dark Factory Workflows
Pre-built pipelines optimized for e-commerce content production. Upload a product image and generate a complete content suite automatically.- Product photography and videos
- UGC-style ads
- Social media kits
- Brand assets
Node System Workflows
Build custom workflows from scratch using 36 node types. Chain AI operations, add conditional logic, and create reusable templates.- Custom image and video pipelines
- Batch processing
- Conditional branching
- Scheduled automation
Workflow Lifecycle
- Create — Build or select a workflow
- Configure — Set parameters, inputs, and options
- Validate — Check for errors and missing required fields
- Execute — Run the workflow and monitor progress
- Review — View, download, and share results
- Iterate — Save as template or re-run with modifications
Building Workflows
Starting Points
From Scratch
Start with an empty canvas. Add nodes, connect them, and configure each step manually.
From Template
Browse 60+ marketplace templates. Clone and customize for your needs.
From Recipe
Use Dark Factory recipes. Pre-configured service combinations for common use cases.
Workflow Components
| Component | Purpose | Example |
|---|---|---|
| Inputs | Starting data | Product image, text prompt, video file |
| Processors | Transform data | Generate image, remove background, upscale |
| Logic | Control flow | If/else conditions, filters, merges |
| Outputs | Final results | Downloadable assets, shared posts, scheduled jobs |
Running Workflows
Execution Modes
Single Run- Process one input through the workflow
- Best for testing and one-off content
- Results appear immediately
- Process multiple inputs simultaneously
- Uses List Node or Dark Factory batch mode
- More efficient credit usage
- Progress tracked per item
Credit Deduction
Credits are deducted as each node executes:- Image generation: 1–4 credits per image
- Video generation: 8–24 credits per video
- Processing: 1–5 credits per operation
- Logic nodes: Free
Execution Status
| Status | Meaning |
|---|---|
| Queued | Waiting for resources |
| Running | Currently processing |
| Complete | Finished successfully |
| Failed | Error occurred |
| Cancelled | User stopped execution |
Automation
Cron Scheduling
Schedule workflows to run automatically:- Daily content generation (0 9 * * *)
- Weekly reports (0 9 * * 1)
- Hourly monitoring (0 * * * *)
Triggers
Future capability (on roadmap): React to events- New product added to Shopify
- Trending content detected
- Competitor ad spotted
Workflow Patterns
Pattern: Generate + Enhance
Pattern: Batch Process
Pattern: Conditional Route
Pattern: Scheduled Content
Managing Workflows
Workspaces
Save workflows as named workspaces:- Auto-saved drafts
- Named templates
- Shared team workflows
- Marketplace submissions
Templates
Save successful workflows as templates:- Name and description
- Default parameters
- Thumbnail preview
- Privacy (private, team, public)
Versioning
Track changes to workflows:- Save versions at milestones
- Compare versions
- Revert to previous version
- See edit history
Best Practices
Performance
Use List Nodes for Batch
Use List Nodes for Batch
Process multiple items in parallel rather than running the workflow multiple times. List nodes are more efficient and cost-effective.
Test Before Scaling
Test Before Scaling
Run workflows on 1–3 items before processing 100. Verify outputs match expectations before committing credits to large batches.
Choose the Right Models
Choose the Right Models
Don’t overpay for quality you don’t need. Flux 2 Pro is excellent but costs more credits. Recraft V3 is cheaper and great for many use cases.
Cache Intermediate Results
Cache Intermediate Results
If a workflow fails mid-way, cached results let you resume from the last successful node instead of starting over.
Organization
Name Workflows Clearly
Name Workflows Clearly
Use descriptive names like “Product Photo to 4K Video” instead of “Workflow_v2_final.”
Group Related Workflows
Group Related Workflows
Add Descriptions
Add Descriptions
Document what the workflow does, expected inputs, and typical outputs. Your future self will thank you.
Use Tags
Use Tags
Troubleshooting
Common Issues
Workflow fails to start
Workflow fails to start
- Check all required fields (marked with *) are filled
- Verify inputs are properly connected
- Ensure sufficient credits are available
- Check for circular dependencies in logic nodes
Results don't match expectations
Results don't match expectations
- Review prompt quality — specific prompts yield better results
- Check model settings — quality tier affects output
- Verify input data quality — AI outputs depend on input quality
- Test individual nodes to isolate the issue
Slow execution
Slow execution
- Video generation takes 2–5 minutes per clip
- Batch processes scale linearly with item count
- Some models are slower than others (noted in model selectors)
- Check if running in parallel or sequence
Credits deducted but no output
Credits deducted but no output
- Check the execution log for error messages
- Some processing failures still consume credits (AI provider costs)
- Contact support with your run ID for credit review