FlowMind AI

FlowMind AI is an AI-first automation builder. Users describe an automation in plain English, Gemini turns it into structured workflow nodes, React Flow renders the workflow, and the backend can save, run, export, or create the workflow in a self-hosted n8n instance.

Features

• Gemini workflow generation with GEMINI_MODEL=gemini-3.1-pro
• React Flow visual workflow canvas
• Conversational editing for generated workflows
• Dashboard metrics, workflow history, and execution logs
• Express API with JWT demo auth
• MongoDB-ready data models
• Local JSON storage fallback for quick portfolio demos
• n8n workflow JSON export and optional n8n API creation

Quick Start

npm install
cp .env.example .env
npm run dev

Open http://localhost:5173.

The app runs without MongoDB or n8n. If GEMINI_API_KEY is missing, it uses a deterministic local generator so the product still works during development.

Environment

PORT=4000
CLIENT_ORIGIN=http://localhost:5173
JWT_SECRET=change-this-secret
GEMINI_API_KEY=your-gemini-api-key
GEMINI_MODEL=gemini-3.1-pro
MONGODB_URI=mongodb://localhost:27017/flowmind-ai
N8N_BASE_URL=http://localhost:5678
N8N_API_KEY=your-n8n-api-key

n8n Local Setup

docker run -it --rm --name n8n -p 5678:5678 -v n8n_data:/home/node/.n8n n8nio/n8n

Create an API key in n8n, add it to .env, then use the n8n button in FlowMind.

API

• POST /api/auth/demo creates a demo session
• POST /api/ai/workflows/generate generates a workflow from a prompt
• POST /api/ai/workflows/edit edits an existing workflow conversationally
• POST /api/workflows saves a workflow
• POST /api/workflows/:id/run simulates an execution
• POST /api/workflows/:id/n8n creates or previews an n8n workflow
• GET /api/workflows/:id/export exports n8n-compatible JSON

Production Notes

For a production deployment, add real Google/GitHub OAuth handlers, encrypt third-party credentials, replace simulated execution with n8n webhook execution tracking, and configure a managed MongoDB database.