Use it daily
Once your servers are connected, here's the whole job. It's mostly reading a list and clicking Apply.
Your one screen: Health
The Health page lists every problem across every server, worst first. Each line is a Finding — a plain-language problem on a specific server. Open one and you get three things:
- What's wrong, in words.
- The exact fix — the precise commands that will run. Read them.
- An Apply button — if the fix is safe enough to automate (see below).
Click Apply → you watch it run live → HyprBox verifies it worked → the finding disappears from the list. Done.
How risky is a fix? (the colour code)
Every fix declares how careful you should be. The button changes accordingly:
| Tier | Meaning | What you do |
|---|---|---|
| SAFE | No real risk | One click. |
| CONFIRM | A real but reversible change | Click, then confirm. (Most fixes.) |
| DANGEROUS | Could disrupt or lock you out | You type a confirmation. |
| MANUAL | Too risky to automate | No button — HyprBox shows you instructions, you do it by hand. |
You're never one accidental click away from breaking a server. And many fixes protect themselves — e.g. the "disable SSH passwords" fix refuses to run unless you already have an SSH key set up, so it can't lock you out.
Testing on your own machine? Only the demo fix is safe to Apply there. The real fixes run real commands — apply those on servers you mean to change.
The three modules, in practice
- HyprGuard (security) — runs automatically on every connected server. You'll see findings like "firewall inactive" or "security updates pending" with a one-click hardening fix.
- HyprVault (backups) — go to Backups, create a policy (where to back up, how often). HyprBox flags databases with no backup.
- HyprWatch (monitoring) — turn it on per server; HyprBox then offers to install a full Prometheus + Grafana monitoring stack and checks it's running.
A few other pages
- Jobs — every fix that ran, with its full live log. Cancel a running one here.
- Backups — your backup policies and their runs.
- Settings — your account, and the tokens that let servers connect.
- Audit (admins) — a record of who did what (logins, fixes, changes), exportable.
The mental model, one more time
You don't manage HyprBox. You read what it found, decide what to fix, and click. It does the work and proves it worked. That's the loop: Discover → Recommend → Preview → Apply → Verify.
Want the exhaustive version (every option, production, writing your own fixes)? → The full guide.