Supabase free-tier project keeps pausing: inactivity limits vs. moving to Pro with micro compute

I run a low-traffic side project on Supabase's free tier. My understanding is that free-tier projects are automatically paused after a period of inactivity (roughly a week, historically), and the database stops responding until it is restored through the dashboard or API. Before I commit to a paid plan, I want to make sure I am diagnosing the right trade-off rather than guessing.

The workload is a small Postgres database well under the historical ~500 MB free-tier cap, with occasional reads and writes and no strict uptime requirement. The concern is that auto-pausing may break things I have not fully accounted for: scheduled jobs, cron triggers, and any health checks or first-request latency when a real user arrives after a quiet stretch.

The documented alternative appears to be the Pro plan with the smallest (micro) compute add-on to keep the project always-on, with the option to scale compute down later. I also understand disk can generally be expanded but not shrunk, so I want to avoid over-provisioning if I do upgrade.

Given pricing and pause thresholds change over time, I plan to verify current numbers on the pricing page first.

1. For a workload like this, is the inactivity pause purely an availability concern, or does it also affect scheduled jobs and cron-based features in ways that make the free tier unsuitable?
2. Is micro compute on Pro the correct minimal always-on configuration, and are there downgrade constraints I should check before resizing?
3. What is the most reliable way to confirm from dashboard logs or usage metrics that pausing is actually what users are hitting?