| name | create-workflow-stub |
| description | Use when the user asks to create a new n8n workflow, scaffold a workflow template, or stub out automation logic. |
Create n8n Workflow Stub
Generate a new n8n workflow template based on requirements.
Usage
/create-workflow-stub [workflow-name]
Use this to quickly scaffold any of the 5 needed workflows (WF1-WF5) or custom workflows.
Instructions
You are a workflow scaffolding specialist. Your job is to create properly structured n8n workflow templates.
Step 1: Determine Workflow Type
If user provides a workflow name, use it. Otherwise, show the 5 planned workflows:
Which workflow would you like to create?
**Planned Workflows (from project memory):**
1. WF1: Research Pipeline (POST /webhook/bomb-research-product)
2. WF2: Ad Copy Generator (POST /webhook/bomb-generate-ads)
3. WF3: Image Generation Pipeline (POST /webhook/bomb-generate-images)
4. WF4: Ad Concept (POST /webhook/bomb-ad-concept)
5. WF5: Ad Clone (POST /webhook/bomb-ad-clone)
6. Custom workflow (specify name)
Enter number or custom name:
Step 2: Gather Requirements
For the selected workflow, ask targeted questions:
- Input: What data triggers this workflow?
- Processing: What transformations/API calls needed?
- Output: Where does the result go? (Airtable, file, webhook response)
- Credentials: Which APIs will be used?
Step 3: Generate Workflow Structure
Create a valid n8n workflow JSON with:
Core Structure:
{
"name": "[Workflow Name]",
"nodes": [
{
"parameters": {
"path": "[webhook-path]",
"responseMode": "responseNode",
"options": {}
},
"id": "webhook-trigger",
"name": "Webhook",
"type": "n8n-nodes-base.webhook",
"typeVersion": 1.1,
"position": [250, 300]
},
{
"parameters": {},
"id": "placeholder-processing",
"name": "Processing Node",
"type": "n8n-nodes-base.noOp",
"typeVersion": 1,
"position": [450, 300],
"notes": "TODO: Add processing logic here"
},
{
"parameters": {
"respondWith": "json",
"responseBody": "={{ $json }}"
},
"id": "webhook-response",
"name": "Response",
"type": "n8n-nodes-base.respondToWebhook",
"typeVersion": 1,
"position": [650, 300]
}
],
"connections": {
"Webhook": {
"main": [[{"node": "Processing Node", "type": "main", "index": 0}]]
},
"Processing Node": {
"main": [[{"node": "Response", "type": "main", "index": 0}]]
}
},
"active": false,
"settings": {},
"versionId": "00000000-0000-0000-0000-000000000000",
"id": "temp-id",
"meta": {},
"tags": []
}
Step 4: Add Workflow-Specific Nodes
Based on workflow type, add appropriate nodes:
For Research Pipeline (WF1):
- Webhook trigger
- HTTP Request to scraping APIs
- Code node for data processing
- Airtable node (create records)
- Response node
For Ad Copy Generator (WF2):
- Webhook trigger
- Airtable node (read product data)
- OpenAI node (generate copy)
- Code node (format output)
- Airtable node (save ad copy)
- Response node
For Image Generation (WF3):
- Webhook trigger
- fal.ai node (generate images)
- AWS S3 node (store images)
- Airtable node (save URLs)
- Response node
Step 5: Save Workflow File
- Convert workflow name to filename: "Research Pipeline" → "research-pipeline.json"
- Save to
/workflows directory
- Create companion markdown file with documentation
Step 6: Generate Documentation
Create [workflow-name].md with:
# [Workflow Name]
## Overview
[Description of what this workflow does]
## Webhook Endpoint
POST http://localhost:5678/webhook/[path]
## Input Schema
```json
{
"productUrl": "string",
"additionalData": "object"
}
Processing Steps
- Step 1 description
- Step 2 description
- Step 3 description
Output Schema
{
"success": "boolean",
"data": "object"
}
Required Credentials
TODO
Testing
curl -X POST http://localhost:5678/webhook/[path] \
-H "Content-Type: application/json" \
-d '{"test": "data"}'
### Step 7: Summary & Next Steps
✓ Created workflow stub: [Workflow Name]
Files Created:
- workflows/[workflow-name].json (workflow definition)
- workflows/[workflow-name].md (documentation)
Next Steps:
- Review the TODO items in the documentation
- Run /import-workflow workflows/[workflow-name].json
- Configure credentials in n8n UI
- Replace placeholder nodes with actual logic
- Test with sample data
Credentials Needed:
[List of required API keys]
## Templates by Workflow Type
### Template 1: Simple Webhook → Process → Respond
For quick transformations and responses
### Template 2: Webhook → Database → AI → Database → Respond
For AI-powered workflows with data persistence
### Template 3: Webhook → Multi-API → Aggregate → Respond
For workflows calling multiple external services
### Template 4: Scheduled Trigger → Process → Store
For automated background jobs
## Notes
- All workflows start inactive by default
- Webhook paths should follow pattern: `/webhook/bomb-[action]`
- Use NoOp nodes as placeholders for incomplete logic
- Include error handling nodes in production workflows