Skip to content

README.md

Latest commit

 

History

History
105 lines (75 loc) · 3.1 KB

README.md

File metadata and controls

105 lines (75 loc) · 3.1 KB

Deployment Guide

Three practical ways to get Lab Manager in front of a user, from fastest evaluation to a real deployment.

Validate The Release Candidate First

Before deploying a new release candidate from a checkout, verify the maintained release gate:

uv sync --dev --frozen
docker compose --env-file .env.example config -q
uv run pytest tests --ignore=tests/bdd -q
bash scripts/run_release_gate.sh

This is the current release standard for the default shipped product surface.

Option 0: Local Trial On Your Laptop

Best for product evaluation, demos, and first-time users who want to try the app before adopting it.

cd lab-manager
bash scripts/bootstrap_local_env.sh "My Lab"
docker compose up -d --build

Then open http://localhost and complete the setup wizard in the browser.

This path is intentionally local-friendly:

  • it generates secrets automatically
  • it sets SECURE_COOKIES=false for localhost HTTP
  • it does not require an AI API key to explore the core product

Option 1: One-Command Install

Run on any Ubuntu/Debian machine (local server, VM, cloud instance):

gh repo clone labclaw/lab-manager
cd lab-manager
bash deploy/install.sh

The installer will:

  • Install Docker if needed
  • Generate all secrets automatically
  • Ask for your lab name, domain, and optional Gemini API key
  • Start all services and verify they're healthy

Requirements: Ubuntu/Debian, 4GB+ RAM, root or sudo access.

This repo is currently private. If you do not use gh repo clone, run the installer from an existing checked-out copy of the repository.

Option 2: DigitalOcean One-Click

Create a fully configured cloud server with one command from your local checked-out repo:

gh repo clone labclaw/lab-manager
cd lab-manager
# Install doctl: https://docs.digitalocean.com/reference/doctl/how-to/install/
doctl auth init
bash deploy/digitalocean/create-droplet.sh --lab-name "My Lab"

This creates a $24/mo droplet (4GB RAM, 2 vCPUs) with:

  • Lab Manager running and accessible via HTTP
  • UFW firewall (ports 22, 80, 443 only)
  • Daily database backups at 2 AM
  • All secrets auto-generated
  • The current local repo snapshot uploaded directly to the droplet

Options:

bash deploy/digitalocean/create-droplet.sh --region sfo3 --name my-lab --lab-name "My Lab"
bash deploy/digitalocean/create-droplet.sh --gemini-api-key "$GEMINI_API_KEY"

Option 3: Manual Docker Compose

For custom setups, see DEPLOY.md in the project root.

cp .env.example .env
# Edit .env with your passwords and settings
docker compose up -d

After Deployment

All three options end the same way:

  1. Open the URL in your browser (the installer prints it)
  2. You'll see a setup wizard -- enter your lab name and create an admin account
  3. Start using Lab Manager immediately

The SQLAdmin panel is available at /admin/ using the generated ADMIN_PASSWORD (printed by the installer or stored in /opt/labclaw/.admin_password on DigitalOcean).

To enable AI features later, add a Gemini API key to .env and restart:

# Edit .env and set GEMINI_API_KEY=your-key
docker compose up -d