MCP N8n Webhook

🚀 MCP N8N Webhook Server

The MCP (Model Context Protocol) server sends data to an n8n webhook for storage and indexing.

🚀 Quick Start

The MCP server enables AI assistants to save structured information in n8n via webhooks. It's suitable for:

• Saving conversations and development logs
• Storing project documentation
• Indexing information for later searches
• Creating content embeddings for RAG

📦 Installation

npm install
npm run build

📚 Documentation

Configuration

Create a .env file based on config.example.env:

cp config.example.env .env

Configure the environment variables:

• N8N_WEBHOOK_URL: The base URL of the n8n webhook
• N8N_WEBHOOK_ID: The unique ID of the webhook
• N8N_API_KEY: The API key for authentication
• N8N_INDEX_NAME: The default index name (optional, default: PROJECT_CONVERSATIONS)
• N8N_TABLE_CATEGORY: The default table category (optional, default: DEVELOPMENT_LOGS)

Usage

Development Mode

npm run dev

Inspector Mode (for testing)

npm run inspector

Production Mode

npm start

Available Tools

n8n_save_data

Save structured information to the n8n webhook.

Parameters:

• projectName (required): The name of the project
• projectCategory (required): The category of the project
• contentType (required): The type of content
• title (required): The title of the content
• description (required): The description of the content
• text (required): The full text
• tags (required): An array of tags
• projectId (optional): The ID of the project
• projectStatus (optional): The status of the project
• language (optional): The language of the content
• indexName (optional): The name of the index
• tableCategory (optional): The table category
• source (optional): The source of the data

n8n_health_check

Verify the connectivity to the n8n webhook.

Payload Structure

The server sends data in the following format:

{
  "metadata": {
    "timestamp": "2025-07-02T10:30:00Z",
    "source": "mcp_assistant",
    "version": "1.0"
  },
  "project": {
    "id": "proj_001",
    "name": "Sistema de Embeddings",
    "category": "AI_ML",
    "status": "ACTIVE"
  },
  "content": {
    "type": "CONVERSATION",
    "title": "Implementación de almacenamiento vectorial",
    "description": "Discusión sobre la arquitectura...",
    "text": "Se está implementando un sistema...",
    "tags": ["embeddings", "vectorial", "conversaciones"],
    "language": "es"
  },
  "indexing": {
    "index_name": "PROJECT_CONVERSATIONS",
    "table_category": "DEVELOPMENT_LOGS"
  }
}

📄 License

MIT