--- name: nocode-orchestrator description: Build and manage automations across Make.com, n8n, Zapier, and Pipedream — onboarding, support tickets, content approval, invoice processing. Use when building and manage automations across make.com, n8n, zapier, and pipedream. domain: productivity author: mahipal license: Apache-2.0 subdomain: productivity tags: - api - nocode - orchestrator - productivity - time-management - tools - voice version: 1.0.0 --- ## Overview Design, build, and manage business process automations using no-code/low-code platforms. Maps manual processes to automated workflows, selects the right platform for each use case, and deploys production-grade automations that businesses pay $2K-$10K per workflow for. A sellable service with low build effort and high leverage across all domains. ## Required Tools - Make.com account (best for complex multi-step workflows) - n8n instance (self-hosted, best for cost-sensitive clients) - Zapier account (best for simple integrations, largest app ecosystem) - Pipedream account (best for code-in-the-middle workflows) - Node.js/Python for custom API connectors - `curl` for API testing and webhook debugging ## Capabilities - Map manual business processes to automated workflows - Select optimal platform based on complexity, cost, and integration needs - Build multi-step workflows with branching, error handling, and retries - Deploy and monitor workflows in production - Optimize costs by routing simple tasks to cheaper platforms - Create reusable workflow templates for common business patterns - Debug failing workflows with structured logging ## When to Use **Trigger phrases:** - "nocode orchestrator" - ""Automate our customer onboarding process"" - ""When a support ticket comes in, triage it and assign to the right person"" - ""Process invoices automatically when they arrive by email"" - "Automate our customer onboarding process" - "When a support ticket comes in, triage it and assign to the right person" - "Process invoices automatically when they arrive by email" - "Sync data between our CRM, email tool, and project management" - "Build a content approval workflow for our team" - Client says: "We do this manually every day, it takes 2 hours" ## When NOT to Use - Task is outside your authorization scope - You need to implement controls (use implementing-* skills) - Task is about analysis, not action (use analyzing-* skills) - You don't have access to target systems - Task requires compliance expertise (consult professionals) - Task is about defense, not offense (use defensive skills) ## Pseudo Code ```python # Example workflow for this skill def execute(input_data): # Step 1: Validate input if not input_data: raise ValueError("Input data is required") # Step 2: Process core logic result = process(input_data) # Step 3: Validate output validate_output(result) return result ``` ### Process Mapping ```markdown ## Process Trigger: New customer signs up (Stripe webhook) Steps: 1. Create customer record in CRM (HubSpot) 2. Send welcome email with setup guide 3. Create onboarding project in PM tool (Linear/Asana) 4. Add to customer Slack channel 5. Schedule Day 3 check-in email 6. Schedule Day 7 follow-up task 7. If enterprise: assign account manager Error handling: - CRM API down → retry 3x, then queue for manual - Email fails → retry, fallback to SMS - Slack channel exists → skip creation, just add member ``` ### Platform Selection ```python def select_platform(process): """Choose optimal platform based on process characteristics.""" criteria = { 'step_count': len(process['steps']), 'has_custom_logic': any(s.get('type') == 'code' for s in process['steps']), 'integration_count': len(set(s['app'] for s in process['steps'])), 'runs_per_month': process['frequency'] * 30, 'budget': process.get('budget', 'medium'), } # Decision matrix if criteria['has_custom_logic'] and criteria['step_count'] > 10: return 'pipedream' # Code flexibility for complex workflows elif criteria['budget'] == 'low' and criteria['runs_per_month'] > 10000: return 'n8n' # Self-hosted, no per-run cost elif criteria['step_count'] <= 5 and criteria['integration_count'] <= 3: return 'zapier' # Simple, fast to build else: return 'make' # Best balance for complex multi-step ``` ### Make.com Workflow Build ```json { "name": "Customer Onboarding", "modules": [ { "id": 1, "app": "webhook", "action": "custom-webhook", "trigger": true, "config": { "url": "https://hook.make.com/abc123", "method": "POST" } }, { "id": 2, "app": "hubspot", "action": "create-contact", "config": { "email": "{{1.body.email}}", "firstname": "{{1.body.name}}", "lifecyclestage": "customer" } }, { "id": 3, "app": "gmail", "action": "send-email", "config": { "to": "{{1.body.email}}", "subject": "Welcome aboard!", "body": "{{templates.welcome_email}}" } }, { "id": 4, "app": "slack", "action": "add-to-channel", "config": { "channel": "#customers", "email": "{{1.body.email}}" }, "error_handler": { "type": "continue", "condition": "already_in_channel" } }, { "id": 5, "app": "router", "condition": "{{1.body.plan}} == 'enterprise'", "routes": [ {"target": 6, "label": "Enterprise path"}, {"target": 7, "label": "Standard path"} ] }, { "id": 6, "app": "asana", "action": "create-task", "config": { "name": "Enterprise onboarding: {{1.body.company}}", "assignee": "account_manager", "due_date": "+7d" } } ] } ``` ### n8n Self-Hosted Setup ```bash # Deploy n8n with Docker docker run -d \ --name n8n \ -p 5678:5678 \ -v ~/.n8n:/home/node/.n8n \ -e N8N_BASIC_AUTH_ACTIVE=true \ -e N8N_BASIC_AUTH_USER=admin \ -e N8N_BASIC_AUTH_PASSWORD=secure_password \ n8nio/n8n # Import workflow via API curl -X POST http://localhost:5678/api/v1/workflows \ -H "Content-Type: application/json" \ -d @workflow.json ``` ### Pipedream Code Step ```javascript // Pipedream Node.js step — custom logic in the middle of a workflow export default defineComponent({ async run({ steps, $ }) { const customer = steps.trigger.event.body; // Custom business logic const tier = customer.mrr > 1000 ? 'enterprise' : customer.mrr > 100 ? 'growth' : 'starter'; // Enrich with external data const companyInfo = await fetch( `https://api.clearbit.com/v2/companies/find?domain=${customer.domain}`, { headers: { Authorization: `Bearer ${process.env.CLEARBIT_KEY}` } } ).then(r => r.json()); return { ...customer, tier, company_size: companyInfo.metrics.employees, industry: companyInfo.category.industry, }; }, }); ``` ### Cost Optimization ```python def optimize_costs(workflows): """Route workflows to cheapest viable platform.""" costs = { 'zapier': {'per_run': 0.025, 'included': 750}, 'make': {'per_run': 0.004, 'included': 10000}, 'n8n': {'per_run': 0, 'monthly_hosting': 20}, 'pipedream': {'per_run': 0.002, 'included': 100000}, } for wf in workflows: monthly_runs = wf['frequency'] * 30 best = min(costs.items(), key=lambda x: max(0, monthly_runs - x[1].get('included', 0)) * x[1].get('per_run', 0) + x[1].get('monthly_hosting', 0) ) wf['recommended_platform'] = best[0] return workflows ``` ### Monitoring Dashboard ```python def check_workflow_health(platform, workflow_id): """Monitor workflow execution status.""" if platform == 'make': resp = httpx.get( f"https://api.make.com/v2/scenarios/{workflow_id}/executions", headers={"Authorization": f"Token {MAKE_API_KEY}"}, params={"limit": 100} ) executions = resp.json()['executions'] failed = [e for e in executions if e['status'] == 'error'] return { 'total': len(executions), 'failed': len(failed), 'success_rate': (len(executions) - len(failed)) / len(executions) * 100, 'last_error': failed[0] if failed else None, } ``` ## Error Handling | Error | Cause | Recovery | |-------|-------|----------| | Webhook timeout | Upstream service slow | Increase timeout, add retry with backoff | | API rate limit | Too many calls | Batch requests, add delays between steps | | Auth token expired | OAuth refresh failed | Re-authenticate, alert user | | Step timeout | Complex logic too slow | Split into sub-workflows, optimize code steps | | Duplicate execution | Webhook fired twice | Add idempotency key, deduplicate in first step | | Platform outage | Make/Zapier down | Queue triggers, replay on recovery | ## Common Patterns - **Fan-out/fan-in**: Trigger N parallel steps, wait for all to complete, merge results - **Error boundary**: Wrap risky steps in error handlers, continue on failure - **Idempotency**: Use unique IDs to prevent duplicate processing on webhook retries - **Rate limiting**: Add deliberate delays between API calls to respect limits - **Dead letter queue**: Failed executions go to a review queue instead of dropping - **Template library**: Build reusable workflow templates for common patterns (onboarding, support triage, invoice processing) ## How to Use 1. Invoke the skill when relevant domain keywords appear in the request 2. Provide required inputs as specified in the skill definition 3. Review the output for correctness before delivering to the user 4. Combine with related skills for complex multi-step workflows ## Verification After completing this skill, confirm: - [ ] Output meets the defined quality and completeness requirements - [ ] All prerequisites are verified and documented - [ ] Error handling covers edge cases - [ ] Results are accurate and actionable ## Anti-Rationalization Table | Rationalization | Reality | |---|---| | "I am too busy to organize" | Disorganization costs more time than organizing. Invest upfront. | | "Multitasking is productive" | Context switching costs 25 minutes per switch. Focus on one thing. | | "I will remember this" | You will not. Write it down. Externalize your memory. |