Initial Setup & Cron Jobs

Applies regardless of whether you use Apache or Docker.

Initial Setup

Open your subdomain in a browser (e.g. https://legal.yourdomain.com). The Paragrafy setup wizard starts automatically and creates the database, admin password, and a random cron secret.

Setting Up Cron Jobs (recommended)

Four endpoints should be called regularly from the outside so scheduled publications go live, backups are created, and webhooks get delivered. All four are protected by a secret key (query parameter ?secret=...), which you’ll find pre-assembled under Settings → Automation (Cron) — you can regenerate it there if needed.

# Publish scheduled changes (every minute, across all projects)
* * * * * curl -fsS "https://legal.yourdomain.com/api/cron/publish?secret=YOUR_CRON_SECRET" > /dev/null

# Process the webhook queue (every 5 minutes)
*/5 * * * * curl -fsS "https://legal.yourdomain.com/api/cron/webhooks?secret=YOUR_CRON_SECRET" > /dev/null

# Daily rolling backup (7 days)
0 3 * * * curl -fsS "https://legal.yourdomain.com/api/cron/backup?secret=YOUR_CRON_SECRET" > /dev/null

# Email report if legal texts are overdue for review (daily)
0 8 * * * curl -fsS "https://legal.yourdomain.com/api/cron/audit?secret=YOUR_CRON_SECRET" > /dev/null

Alternatively, an external uptime monitor (e.g. Uptime Kuma, healthchecks.io) can serve as a “cron” that calls these URLs at the desired interval.

Paragrafy remains usable even without cron set up: scheduled publications are also checked automatically whenever someone visits the respective project domain (zero-config fallback) — but with very little traffic, this can delay going live. Backup and the webhook queue can always be triggered manually from settings; an endpoint called without a secret or with the wrong one responds with HTTP 403.

Next, continue with Webhooks, which are delivered via the /api/cron/webhooks job.