The xyz.sh wrapper adds no telemetry and does not perform hidden actions. You can read the full source on the page or with curl -fsSL xyz.sh/ .sh before running it. Piping a remote script to bash is still a risk — review the source first. The default path runs the third-party open-source YABS project (masonr/yet-another-bench-script, MIT). It may download bench binaries and send traffic to public iperf servers.

Which operating systems are supported?

Declared support: Debian, Ubuntu, CentOS, Rocky Linux, and AlmaLinux. The script detects the distribution and asks before continuing on an unrecognized OS. Syntax was checked with bash -n and ShellCheck; installer scripts have not been end-to-end verified on every listed distribution.

No. It can run without root, though some fields or tests may be limited.

How do I uninstall it?

There is nothing to uninstall from the wrapper. --local leaves no files. If you ran YABS, check any temporary download paths it printed.

What scores should I expect?

This site does not publish or invent scores or machine comparisons. Results appear only in your terminal and vary with time, load, and network conditions.

VPS Benchmark Tutorial (YABS Wrapper, No Fabricated Scores)

This tutorial shows how to benchmark a VPS with xyz.sh/bench. The default path wraps open-source YABS; --local is a rough test that downloads no third-party tools. This page does not publish or invent scores.

Two modes

Default (YABS): curl -fsSL xyz.sh/bench | bash previews the script, then runs https://github.com/masonr/yet-another-bench-script after confirmation. YABS may download fio, iperf3, and Geekbench, and may send traffic to public iperf servers. Flags such as -f or -g are forwarded to YABS (for example to skip fio or Geekbench).

Local rough test: curl -fsSL xyz.sh/bench | bash -s -- --local runs an awk loop and a 1GiB dd only, with no third-party download. This is not a standard benchmark; it is only a quick sanity check.

About results

Results print only in your terminal. They vary with time, CPU load, and network conditions, and should not be treated as a fixed machine score or mixed with unofficial comparison tables.

The wrapper itself installs no persistent software. --local uses a temp file and removes it. YABS may download fio, iperf3, or Geekbench into a temporary directory and typically does not change system config; clean up any leftover paths YABS reported if needed.