Files
n8n-mcp/README.md
czlonkowski 88dd66bb7a refactor: make optimized Dockerfile the default - remove full variant
- Rename Dockerfile.optimized to Dockerfile (now the default)
- Keep old Dockerfile as Dockerfile.old for reference
- Update GitHub Actions to use default Dockerfile
- Remove build-full job - only one image variant now
- Remove docker-compose.optimized.yml and other variants
- Update all documentation to reflect single image approach

The optimized 283MB image is now the only n8n-MCP Docker image.
This simplifies the user experience and provides the best solution
for all use cases.

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <noreply@anthropic.com>
2025-06-14 14:00:36 +02:00

320 lines
7.9 KiB
Markdown

# n8n-MCP
A Model Context Protocol (MCP) server that provides AI assistants with comprehensive access to n8n node documentation, properties, and operations.
## Overview
n8n-MCP serves as a bridge between n8n's workflow automation platform and AI models, enabling them to understand and work with n8n nodes effectively. It provides structured access to:
- 📚 **458 n8n nodes** from both n8n-nodes-base and @n8n/n8n-nodes-langchain
- 🔧 **Node properties** - 98.7% coverage with detailed schemas
-**Node operations** - 57.9% coverage of available actions
- 📄 **Documentation** - 88.6% coverage from official n8n docs
- 🤖 **AI tools** - 35 AI-capable nodes detected
## Features
- **Comprehensive Node Information**: Access properties, operations, credentials, and documentation for all n8n nodes
- **AI Tool Detection**: Automatically identifies nodes with AI capabilities (usableAsTool)
- **Versioned Node Support**: Handles complex versioned nodes like HTTPRequest and Code
- **Fast Search**: SQLite with FTS5 for instant full-text search across all documentation
- **MCP Protocol**: Standard interface for AI assistants to query n8n knowledge
## Quick Start
### Prerequisites
- Node.js (any version - automatic fallback to pure JavaScript if needed)
- npm or yarn
- Git
> **Note**: The project uses an intelligent database adapter that automatically falls back to a pure JavaScript implementation (sql.js) if native dependencies fail to load. This ensures compatibility with any Node.js version, including Claude Desktop's bundled runtime.
### Installation
1. Clone the repository:
```bash
git clone https://github.com/yourusername/n8n-mcp.git
cd n8n-mcp
```
2. Clone n8n documentation (required for full documentation coverage):
```bash
git clone https://github.com/n8n-io/n8n-docs.git ../n8n-docs
```
3. Install dependencies:
```bash
npm install
```
4. Build the project:
```bash
npm run build
```
5. Initialize the database:
```bash
npm run rebuild
```
6. Validate the installation:
```bash
npm run test-nodes
```
## Docker Quick Start 🐳
The easiest way to get started is using Docker:
### Option 1: Simple HTTP Server (Recommended)
1. Create a `.env` file:
```bash
# Generate a secure token
AUTH_TOKEN=$(openssl rand -base64 32)
echo "AUTH_TOKEN=$AUTH_TOKEN" > .env
```
2. Run with Docker Compose:
```bash
docker compose up -d
```
3. Test the server:
```bash
curl http://localhost:3000/health
```
4. Configure Claude Desktop with mcp-remote:
```json
{
"mcpServers": {
"n8n-remote": {
"command": "npx",
"args": [
"-y",
"@modelcontextprotocol/mcp-remote@latest",
"connect",
"http://localhost:3000/mcp"
],
"env": {
"MCP_AUTH_TOKEN": "your-auth-token-here"
}
}
}
}
```
### Option 2: Local stdio Mode (Direct Docker)
```json
{
"mcpServers": {
"n8n-docker": {
"command": "docker",
"args": [
"run",
"--rm",
"-i",
"-e", "MCP_MODE=stdio",
"-v", "n8n-mcp-data:/app/data",
"ghcr.io/czlonkowski/n8n-mcp:latest"
]
}
}
}
```
### Building Locally
To build the Docker image locally:
```bash
docker build -t n8n-mcp:local .
```
For detailed Docker documentation, see [Docker Deployment Guide](./docs/DOCKER_README.md).
## Usage
### With Claude Desktop
1. Edit your Claude Desktop configuration:
- macOS: `~/Library/Application Support/Claude/claude_desktop_config.json`
- Windows: `%APPDATA%\Claude\claude_desktop_config.json`
- Linux: `~/.config/Claude/claude_desktop_config.json`
2. Add the n8n-documentation server:
```json
{
"mcpServers": {
"n8n-documentation": {
"command": "node",
"args": [
"/path/to/n8n-mcp/dist/mcp/index.js"
]
}
}
}
```
3. Restart Claude Desktop
> **Note**: The Node.js wrapper script (`mcp-server-v20.sh`) is no longer required! The project now automatically handles version mismatches.
### Available MCP Tools
- `list_nodes` - List all n8n nodes with filtering options
- `get_node_info` - Get detailed information about a specific node
- `search_nodes` - Full-text search across all node documentation
- `list_ai_tools` - List all AI-capable nodes
- `get_node_documentation` - Get parsed documentation for a node
- `get_database_statistics` - View database metrics and coverage
### Example Queries
```typescript
// List all trigger nodes
list_nodes({ isTrigger: true })
// Get info about the HTTP Request node
get_node_info({ nodeType: "nodes-base.httpRequest" })
// Search for OAuth-related nodes
search_nodes({ query: "oauth" })
// Find AI-capable tools
list_ai_tools()
```
## Development
### Commands
```bash
npm run build # Build TypeScript
npm run rebuild # Rebuild node database
npm run test-nodes # Test critical nodes
npm run validate # Validate node data
npm start # Start MCP server
npm test # Run tests
npm run typecheck # Check TypeScript types
```
### Architecture
```
src/
├── loaders/ # Node package loaders
├── parsers/ # Node metadata parsers
├── mappers/ # Documentation mappers
├── database/ # SQLite repository
├── scripts/ # Build and test scripts
└── mcp/ # MCP server implementation
```
### Node.js Version Compatibility
The project works with any Node.js version thanks to automatic adapter fallback:
- **Primary**: Uses `better-sqlite3` when compatible (faster)
- **Fallback**: Uses `sql.js` when version mismatch detected (pure JS)
- **Automatic**: No manual configuration needed
## Technical Architecture
### Database Adapter
The project features an intelligent database adapter that ensures compatibility across different Node.js versions:
1. **Primary**: Attempts to use `better-sqlite3` for optimal performance
2. **Fallback**: Automatically switches to `sql.js` (pure JavaScript) if:
- Native modules fail to load
- Node.js version mismatch is detected
- Running in restricted environments (like Claude Desktop)
This dual-adapter approach means:
- ✅ Works with any Node.js version
- ✅ No compilation required in fallback mode
- ✅ Maintains full functionality with either adapter
- ✅ Automatic persistence with sql.js
### Performance Characteristics
- **better-sqlite3**: Native performance, ~10-50x faster
- **sql.js**: Pure JavaScript, ~2-5x slower but still responsive
- Both adapters support the same API for seamless operation
## Metrics
Current implementation achieves:
- ✅ 458/458 nodes loaded (100%)
- ✅ 452 nodes with properties (98.7%)
- ✅ 265 nodes with operations (57.9%)
- ✅ 406 nodes with documentation (88.6%)
- ✅ 35 AI-capable tools detected
- ✅ All critical nodes validated
## Remote Deployment
### HTTP Server Mode
n8n-MCP now supports HTTP mode for remote deployments. This allows you to:
- Host the MCP server on a cloud VPS
- Connect from Claude Desktop using mcp-remote
- Single-user design for private use
- Simple token-based authentication
### Quick Start for HTTP Mode
1. On your server:
```bash
# Set environment variables
export MCP_MODE=http
export AUTH_TOKEN=$(openssl rand -base64 32)
# Start the server
npm run start:http
```
2. On your client, configure Claude Desktop with mcp-remote:
```json
{
"mcpServers": {
"n8n-remote": {
"command": "npx",
"args": [
"-y",
"@modelcontextprotocol/mcp-remote@latest",
"connect",
"https://your-server.com/mcp"
],
"env": {
"MCP_AUTH_TOKEN": "your-auth-token"
}
}
}
}
```
For detailed instructions, see [HTTP Deployment Guide](./docs/HTTP_DEPLOYMENT.md).
## Contributing
1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Run tests and validation
5. Submit a pull request
## License
This project uses the Sustainable Use License. See LICENSE file for details.
Copyright (c) 2024 AiAdvisors Romuald Czlonkowski
## Acknowledgments
- n8n team for the excellent workflow automation platform
- Anthropic for the Model Context Protocol specification