14 KiB
14 KiB
Create n8n Workflow - Workflow Instructions
<critical>The workflow execution engine is governed by: {project_root}/{bmad_folder}/core/tasks/workflow.xml</critical>
<critical>You MUST have already loaded and processed: {installed_path}/workflow.yaml</critical>
<critical>This workflow creates a new n8n workflow from scratch based on user requirements.</critical>
<workflow>
<step n="0" goal="Contextual Analysis (Smart Elicitation)">
<critical>Before asking any questions, analyze what the user has already told you</critical>
<action>Review the user's initial request and conversation history</action>
<action>Extract any mentioned: workflow type, trigger, integrations, complexity, requirements</action>
<check if="ALL requirements are clear from context">
<action>Summarize your understanding</action>
<action>Skip directly to Step 2 (Check Context7 MCP)</action>
</check>
<check if="SOME requirements are clear">
<action>Note what you already know</action>
<action>Only ask about missing information in Step 1</action>
</check>
<check if="requirements are unclear or minimal">
<action>Proceed with full elicitation in Step 1</action>
</check>
</step>
<step n="1" goal="Gather Requirements" elicit="true">
<action>Ask Question 1: "What type of automation workflow do you need?"</action>
<action>Present numbered options:
1. Webhook-based - Receive HTTP requests and process data
2. Scheduled - Run on a schedule (cron/interval)
3. Event-driven - Trigger from external service events
4. Manual - Execute on demand
5. Database-driven - Trigger from database changes
6. Other - Describe your specific trigger needs
</action>
<action>WAIT for user selection (1-6)</action>
<action>Store selection in {{workflow_type}}</action>
<check if="selection is 6 (Other)">
<action>Ask: "Please describe your trigger requirements"</action>
<action>WAIT for user input</action>
<action>Store in {{workflow_type}}</action>
</check>
<action>Ask Question 2: "What integrations or services will this workflow use?"</action>
<action>Present numbered options:
1. HTTP/REST APIs - Generic API calls
2. Databases - PostgreSQL, MySQL, MongoDB, etc.
3. Cloud Services - Google Sheets, Slack, Notion, Airtable
4. Email - Send/receive emails
5. CRM - HubSpot, Salesforce, etc.
6. Custom Code - JavaScript/Python logic
7. Multiple - Combination of above
8. Other - Specify integrations
</action>
<action>WAIT for user selection (1-8)</action>
<action>Store selection in {{integrations}}</action>
<check if="selection is 7 or 8">
<action>Ask: "Please list all integrations needed (comma-separated)"</action>
<action>WAIT for user input</action>
<action>Store in {{integrations}}</action>
</check>
<action>Ask Question 3: "How complex is your workflow logic?"</action>
<action>Present numbered options:
1. Simple (3-5 nodes) - Linear flow with minimal logic
2. Medium (6-10 nodes) - Some conditional logic or branching
3. Complex (11-20 nodes) - Multiple branches, loops, or error handling
4. Very Complex (20+ nodes) - Advanced logic with parallel processing
</action>
<action>WAIT for user selection (1-4)</action>
<action>Store selection in {{complexity}}</action>
<action>Ask Question 4: "Do you need error handling and retry logic?"</action>
<action>Present numbered options:
1. No - Simple workflow without error handling
2. Basic - Retry failed operations automatically
3. Advanced - Custom error handling with notifications
4. Comprehensive - Full error workflow with logging and alerts
</action>
<action>WAIT for user selection (1-4)</action>
<action>Store selection in {{error_handling}}</action>
<action>Ask Question 5: "What should the workflow be named?"</action>
<action>WAIT for user input</action>
<action>Store in {{workflow_name}}</action>
<action>Ask Question 6: "Where should the workflow file be saved?"</action>
<action>Present numbered options:
1. Default location - workflows/[workflow-name].json
2. Custom path - Specify your own file path
3. Project root - Save in main project directory
4. Specific folder - Choose from existing folders
</action>
<action>WAIT for user selection (1-4)</action>
<check if="selection is 2 or 4">
<action>Ask for specific path</action>
<action>WAIT for user input</action>
</check>
<action>Store final path in {{save_location}}</action>
<action>Ask Question 7: "Any additional requirements or specific logic needed?"</action>
<action>Present numbered options:
1. No - Proceed with current requirements
2. Yes - Describe additional requirements
</action>
<action>WAIT for user selection (1-2)</action>
<check if="selection is 2">
<action>Ask: "Please describe additional requirements"</action>
<action>WAIT for user input</action>
<action>Store in {{additional_requirements}}</action>
</check>
</step>
<step n="2" goal="Check Context7 MCP Availability">
<action>Check if Context7 MCP server is configured</action>
<action>Try to list available MCP tools</action>
<check if="Context7 MCP is available">
<action>Store true in {{context7_available}}</action>
<action>Proceed to Step 3</action>
</check>
<check if="Context7 MCP is NOT available">
<action>Store false in {{context7_available}}</action>
<action>Inform user: "Context7 MCP server is not configured. This provides up-to-date n8n documentation."</action>
<action>Ask: "Would you like help setting up Context7 MCP? (yes/no)"</action>
<action>WAIT for user response</action>
<check if="user says yes">
<action>Guide user through Context7 MCP setup:</action>
<action>1. Inform user: "Context7 MCP requires the 'uvx' command (part of uv Python package manager)"</action>
<action>2. Ask: "Do you have 'uv' installed? (yes/no/not sure)"</action>
<action>WAIT for response</action>
<check if="user says no or not sure">
<action>Provide installation guidance:</action>
<action>- macOS/Linux: curl -LsSf https://astral.sh/uv/install.sh | sh</action>
<action>- Windows: powershell -c "irm https://astral.sh/uv/install.ps1 | iex"</action>
<action>- Or via pip: pip install uv</action>
<action>Ask user to install and confirm when ready</action>
<action>WAIT for confirmation</action>
</check>
<action>3. Inform user: "Please configure Context7 MCP in your IDE's MCP settings with:"</action>
<action>
Server name: context7
Command: uvx
Args: ["context7"]
Auto-approve tools: ["resolve-library-id", "get-library-docs"]
</action>
<action>4. Ask user to reconnect/restart MCP servers in their IDE</action>
<action>5. Wait for user confirmation that MCP is reconnected</action>
<action>6. Verify Context7 is now available</action>
<action>Store result in {{context7_available}}</action>
</check>
<check if="user says no">
<action>Inform: "Proceeding without Context7. Using built-in n8n knowledge."</action>
<action>Proceed to Step 4</action>
</check>
</check>
</step>
<step n="3" goal="Query Context7 for n8n Documentation">
<check if="{{context7_available}} is true">
<action>Resolve n8n library ID using Context7</action>
<action>Query Context7 for relevant n8n documentation based on:</action>
<action>- Workflow type: {{workflow_type}}</action>
<action>- Integrations: {{integrations}}</action>
<action>- Complexity: {{complexity}}</action>
<action>Store relevant documentation snippets for reference</action>
</check>
<check if="{{context7_available}} is false">
<action>Skip Context7 query</action>
<action>Use built-in knowledge and templates</action>
</check>
</step>
<step n="4" goal="Plan Workflow Structure">
<action>Based on gathered requirements, plan the workflow structure:</action>
<action>1. Identify trigger node type</action>
<action>2. List all action nodes needed</action>
<action>3. Identify conditional logic (IF nodes)</action>
<action>4. Plan data transformations (Set/Code nodes)</action>
<action>5. Design error handling strategy</action>
<action>6. Map node connections</action>
<action>Present the planned structure to user in clear format:</action>
<action>- Trigger: [trigger type]</action>
<action>- Nodes: [list of nodes with descriptions]</action>
<action>- Logic: [conditional branches, loops]</action>
<action>- Error Handling: [strategy]</action>
<action>Ask: "Does this structure meet your needs?"</action>
<action>Present numbered options:
1. Yes - Proceed with this structure
2. No - Adjust the structure
3. Add more details - Provide additional requirements
</action>
<action>WAIT for user selection (1-3)</action>
<check if="selection is 2 or 3">
<action>Ask: "What changes or additions are needed?"</action>
<action>WAIT for user input</action>
<action>Adjust structure based on feedback</action>
<action>Repeat this step</action>
</check>
</step>
<step n="5" goal="Load Templates and Resources">
<action>Load {{templates}} file</action>
<action>Identify closest matching template based on workflow type</action>
<action>Load {{helpers}} for node creation guidelines</action>
<action>Extract relevant template sections</action>
</step>
<step n="6" goal="Build Workflow JSON">
<critical>Follow guidelines from {{helpers}} for proper node creation</critical>
<action>Initialize workflow structure:</action>
<substep>
{
"name": "{{workflow_name}}",
"nodes": [],
"connections": {},
"active": false,
"settings": {
"executionOrder": "v1"
},
"tags": []
}
</substep>
<action>Build nodes ONE at a time following these rules:</action>
<substep>For Each Node:
1. Generate unique node ID
2. Set node name (unique, descriptive)
3. Set node type (e.g., n8n-nodes-base.webhook)
4. Set typeVersion (usually 1)
5. Calculate position:
- First node (trigger): [250, 300]
- Subsequent nodes: add 220 to x for each step
- Branches: adjust y by ±100
6. Configure parameters based on node type
7. Add credentials if needed
8. Set error handling options if required:
- continueOnFail: true/false
- retryOnFail: true/false
- maxTries: number
- waitBetweenTries: milliseconds
</substep>
<substep>For Each Connection:
1. Identify source node name
2. Identify target node name
3. Create connection entry in connections object:
"Source Node Name": [
[
{
"node": "Target Node Name",
"type": "main",
"index": 0
}
]
]
4. For IF nodes:
- index 0 = true branch
- index 1 = false branch
5. Validate all referenced nodes exist
</substep>
<substep>Node Building Order:
1. Trigger node
2. Main flow action nodes
3. Conditional nodes (IF/Switch)
4. Branch nodes
5. Merge nodes
6. Final action nodes
7. Connect all nodes
</substep>
<action>Apply error handling based on {{error_handling}} selection</action>
<action>Validate all node IDs are unique</action>
<action>Validate all node names are unique</action>
<action>Validate all connections reference existing nodes</action>
</step>
<step n="7" goal="Optimize and Save">
<action>Review workflow for optimization:</action>
<action>- Remove any unused nodes</action>
<action>- Verify node positions don't overlap</action>
<action>- Ensure proper spacing and alignment</action>
<action>- Validate all required parameters are set</action>
<action>Save workflow to {{save_location}}</action>
</step>
<step n="8" goal="Validate JSON Syntax">
<critical>NEVER delete the file if validation fails - always fix syntax errors</critical>
<action>Run: node -e "JSON.parse(require('fs').readFileSync('{{save_location}}', 'utf8')); console.log('✓ Valid JSON')"</action>
<check if="validation fails (exit code 1)">
<action>Read the error message carefully - it shows the syntax error and position</action>
<action>Open the file and navigate to the error location</action>
<action>Fix the syntax error (add missing comma, bracket, or quote as indicated)</action>
<action>Save the file</action>
<action>Re-run validation with the same command</action>
<action>Repeat until validation passes</action>
</check>
<action>Once validation passes, confirm with user: "n8n workflow created at {{save_location}}"</action>
</step>
<step n="9" goal="Provide Usage Instructions">
<action>Inform user how to use the workflow:</action>
<action>1. Import the JSON file into n8n</action>
<action>2. Configure credentials for integrated services</action>
<action>3. Test the workflow with sample data</action>
<action>4. Activate the workflow when ready</action>
<action>Ask: "Would you like me to explain any part of the workflow?"</action>
<action>Present numbered options:
1. No - I'm good to go
2. Yes - Explain specific nodes
3. Yes - Explain the overall flow
4. Yes - Explain how to test it
</action>
<action>WAIT for user selection (1-4)</action>
<check if="selection is 2, 3, or 4">
<action>Provide requested explanation</action>
</check>
</step>
<step n="10" goal="Validate Content">
<invoke-task>Validate against checklist at {{validation}} using {{bmad_folder}}/core/tasks/validate-workflow.xml</invoke-task>
</step>
</workflow>