n8n Manager

Local development tool that spins up an n8n instance with Docker, imports workflows and credentials organized by module, and auto-logs into the browser.

Note: This tool is designed for local development and testing only. It uses HTTP, default credentials, and SQLite — not suitable for production.

Requirements

• Docker and Docker Compose
• curl
• python3 (for browser auto-login)

Quick start

# 1. Set environment variables (optional)
cp .env.example .env

# 2. Create a fresh instance
./manager.sh init

Commands

./manager.sh init

Full setup from scratch. Destroys the previous instance, creates a new one, imports everything, and opens the browser already authenticated.

./manager.sh update

Imports workflows and credentials into a running instance. Does NOT destroy data or restart the container.

# Add a new module or update an existing one
./manager.sh update

Module structure

Each workflow is a self-contained module inside workflows/:

workflows/
└── my_workflow/
    ├── workflow.json          # Workflow definition
    └── credentials/           # Credentials it needs
        └── my_credential.json

Adding a new module

mkdir -p workflows/my_new_workflow/credentials
# Drop workflow.json and credentials inside
./manager.sh update

The workflows/ directory is mounted as a read-only volume in the container. The script iterates each subfolder, imports root .json files as workflows and credentials/*.json as credentials.

Configuration

Edit .env to change:

Variable	Default	Description
N8N_IMAGE	n8nio/n8n:latest	n8n Docker image
N8N_PORT	5678	Exposed port
N8N_OWNER_EMAIL	admin@example.com	Owner user email
N8N_OWNER_PASSWORD	Changeme123	Owner user password
N8N_OWNER_FIRST_NAME	Admin	Owner first name
N8N_OWNER_LAST_NAME	User	Owner last name

Notes

• init destroys the previous volume and creates a clean instance.
• update only imports — safe to run multiple times (idempotent by ID).
• Shared credentials can exist in multiple modules — n8n deduplicates by ID.
• Imports read directly from the mounted volume, no docker cp needed.