n8n MCP Backup Server
Enables AI assistants to back up, list, and restore n8n workflows and PostgreSQL databases via natural language, using S3-compatible storage for backups.
README
n8n MCP Backup Server
An MCP (Model Context Protocol) server that lets an AI assistant (Claude Desktop) back up, list, and restore n8n workflows and PostgreSQL databases — all through natural language commands.
"Take a backup of my n8n workflows" → Claude calls the MCP tool → backup uploaded to S3.
What it does
| Tool | Description |
|---|---|
take_backup |
Exports all n8n workflows and credentials to a timestamped JSON and uploads to S3 |
list_backups |
Lists available backups in S3, newest first with sizes |
restore_backup |
Downloads a backup from S3 and imports workflows back into n8n |
take_postgres_backup |
Streams a pg_dump directly to S3 via multipart upload (no temp file) |
list_postgres_backups |
Lists available Postgres backups in S3 |
restore_postgres_backup |
Downloads a Postgres backup from S3 and restores via pg_restore |
Architecture
Claude Desktop
│
│ stdio (MCP)
▼
mcp-server (FastMCP)
│
├──► n8n API (export / import workflows)
├──► S3 Storage (store / retrieve backup files)
└──► PostgreSQL (pg_dump / pg_restore)
All services run as Docker containers and communicate over an internal Docker network. S3 storage defaults to Garage (self-hosted) but works with any S3-compatible backend (AWS S3, MinIO, etc.).
Tech Stack
- Python + FastMCP — MCP server with stdio transport
- boto3 — S3-compatible storage client
- requests — n8n REST API calls
- Docker Compose — orchestrates all services
- Garage v1.3.1 — self-hosted S3-compatible object storage
- PostgreSQL 16 — database with backup/restore support
Quick Start
1. Clone the repo
git clone https://github.com/TameemAlkadiki/n8n-mcp-backup.git
cd n8n-mcp-backup
2. Configure environment
cp .env.example .env
Open .env and fill in all required values. See Configuration below.
3. Set up Garage (S3 storage)
Place your garage.toml at ./garage/config/garage.toml.
See the Garage quick-start guide for setup instructions.
4. Start all services
docker compose up -d
5. Initialize Garage buckets
After first boot, create the S3 buckets:
# Apply cluster layout
docker exec garage /garage layout assign -z default -c 1G <NODE_ID>
docker exec garage /garage layout apply --version 1
# Create buckets
docker exec garage /garage bucket create n8n-backup
docker exec garage /garage bucket create postgres-backup
# Create access key and allow bucket access
docker exec garage /garage key create mcp-key
docker exec garage /garage bucket allow n8n-backup --read --write --key mcp-key
docker exec garage /garage bucket allow postgres-backup --read --write --key mcp-key
Copy the generated key ID and secret into your .env as S3_ACCESS_KEY_ID and S3_SECRET_ACCESS_KEY, then restart:
docker compose restart mcp-server
6. Connect Claude Desktop
Add this to your claude_desktop_config.json:
{
"mcpServers": {
"n8n-backup": {
"command": "docker",
"args": ["exec", "-i", "mcp_backup_server", "python", "server.py"]
}
}
}
Restart Claude Desktop. You can now say:
- "Take a backup of my n8n workflows"
- "List available backups"
- "Restore the backup from yesterday"
- "Take a Postgres backup"
Configuration
All configuration is via environment variables in .env. Copy .env.example to get started.
| Variable | Required | Description |
|---|---|---|
N8N_ENCRYPTION_KEY |
✅ | 32-byte hex key for n8n credential encryption |
N8N_API_KEY |
✅ | n8n API key (Settings → n8n API → Create) |
N8N_URL |
optional | n8n base URL (default: http://n8n:5678) |
GARAGE_RPC_SECRET |
✅ | Garage cluster RPC secret |
GARAGE_ADMIN_TOKEN |
✅ | Garage admin API token (for web UI) |
S3_ACCESS_KEY_ID |
✅ | S3 access key ID |
S3_SECRET_ACCESS_KEY |
✅ | S3 secret access key |
S3_ENDPOINT |
optional | S3 endpoint (default: http://garage:3900) |
S3_REGION |
optional | S3 region (default: us-east-1) |
S3_BUCKET |
optional | Bucket for n8n backups (default: n8n-backup) |
S3_BUCKET_POSTGRES |
optional | Bucket for Postgres backups (default: postgres-backup) |
POSTGRES_DB |
✅ | PostgreSQL database name |
POSTGRES_USER |
✅ | PostgreSQL user |
POSTGRES_PASSWORD |
✅ | PostgreSQL password |
PG_HOST |
optional | Postgres host (default: postgres) |
PG_PORT |
optional | Postgres port (default: 5432) |
Project Structure
n8n-mcp-backup/
├── server.py # FastMCP server — all 6 tools
├── docker-compose.yml # All services: n8n, mcp-server, garage, postgres
├── Dockerfile # MCP server container build
├── requirements.txt # Python dependencies
├── .env.example # Environment variable template
└── garage/
└── config/
└── garage.toml # Garage storage config (not included — see setup)
Notes
- Credentials are not restored by
restore_backup— n8n's API does not expose credential secrets on export. After restoring workflows, re-enter credentials manually in n8n Settings → Credentials. - Multipart upload is used for Postgres backups, meaning no temporary files are written to disk — the dump streams directly from
pg_dumpto S3. - The MCP server uses stdio transport — it is spawned per-session by Claude Desktop via
docker exec, not run as a persistent HTTP service.
License
MIT
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.