How bagussundaru/mcp-server-client-wa works
Project type and purpose
• This is an n8n-based integration “tool” (an open-source workflow orchestration project) that connects WhatsApp (via Business API or Twilio) to an AI assistant (OpenRouter), plus optional services (translator, calculator, scheduler), and persists conversation context in a database.
File structure
• Top-level files are JSON exports of n8n workflows:
• ai-agent-wa-ultimate-n8n.json
• mcp_client_workflow.json
• mcp_server_workflow.json
• README.md describes how to import these workflows into an n8n instance, configure credentials, set up webhooks, error catching, and additional features (translation, search, reminders).
• No traditional “src” or “lib” folders since n8n workflows are authored via GUI and stored as JSON.
README for additional insights
• Workflow “MCP WhatsApp Integration” is composed of:
• Trigger nodes (Webhook for incoming WhatsApp messages)
• HTTP Request nodes (to WhatsApp Business API or Twilio)
• AI Assistant node (calls OpenRouter)
• Tools Router node (dispatches to calculator, translator, scheduler, or AI)
• DB nodes (MongoDB/MySQL) to store and retrieve conversation history
• Error Catcher node for graceful error handling
• Credentials setup, ngrok for local tunneling, and production deployment notes (Railway, DigitalOcean).