Set up platform alerting

## Summary
Set up alerting for platform and infrastructure health.

## Scope
- Evaluate whether to use an existing Apache/open-source project or implement a small internal alerting setup.
- Monitor platform_web, platform_api, Caddy, Postgres, MongoDB, Woodpecker CI, disk usage, backup jobs, and critical cron/batch tasks.
- Define alert destinations and severity levels.
- Keep the first version simple and maintainable.

## Acceptance Criteria
- Chosen alerting approach is documented.
- Critical services and backup jobs have alerts.
- Disk usage and service availability are monitored.
- Alert delivery is tested.
- Runbook notes exist for common alerts.

Use the issue scope as the starting map: review platform_web, platform_api, Caddy, Postgres, MongoDB, Woodpecker CI, disk usage, backups, and cron or batch tasks. Document the chosen alerting approach, define destinations and severity levels, add the required critical-service and backup alerts, test delivery, and record common-alert runbook notes.

Written by the indexing model from the issue text.