Getting Started

Getting Started

You only need Docker. A GPU for local AI is optional — a cloud API key works fine, and the app runs with no AI at all (those features just stay idle).

Easiest — run it on your own machine

git clone https://github.com/Paco5687/GlucoPilot.git glucopilot
cd glucopilot
./install.sh

The installer checks Docker, generates your config (with a random secret key), asks a couple of quick questions (port, timezone, an optional AI key), starts the app, and prints the link — usually http://localhost:8000. Open it, create your login, and finish setup on the Settings page. That's the whole thing.

Server deploy (domain + HTTPS)

For a public deployment, HTTPS is required (the OAuth data-source redirects need it). Any reverse proxy works; the repo ships Traefik labels.

git clone https://github.com/Paco5687/GlucoPilot.git glucopilot
cd glucopilot
cp .env.example .env

Edit .env:

• APP_SECRET_KEY — set a long random string (keep it stable).
• APP_PUBLIC_URL — your HTTPS URL.
• Reverse-proxy settings (TRAEFIK_*) or adapt to nginx/Caddy — see docs/DEPLOY.md.

Everything else — integration credentials, AI provider — is entered later in the app's Settings page and stored in the database.

docker compose up -d          # prebuilt image; add --build to build from source

Open your APP_PUBLIC_URL. On first visit you create the admin username and password (stored only as an Argon2id hash).

First steps in the app

1. Settings → AI — pick a provider and paste a key (or point at a local model).
2. Connections — connect your data sources (see Connecting Data Sources).
3. Let the background sync run; the Dashboard and Explorer fill in.

Forgot your password?

The login page has a reset flow, or from the server:

docker compose exec glucopilot python -m server.reset_password

Backups

All state lives in the glucopilot_data Docker volume:

docker run --rm -v glucopilot_data:/data -v "$PWD":/backup alpine \
  tar czf /backup/glucopilot-backup.tar.gz -C /data .