Skip to content

Security: rtmongold/linkprobe

Security

SECURITY.md

Security Policy

Supported versions

Security fixes are applied to the latest release on main. Older tags may not receive backports.

Version Supported
0.2.0 yes
0.1.0 and older no

Reporting a vulnerability

Do not open public GitHub issues for security problems.

Report privately to rtmongold@gmail.com with:

  • linkprobe version (linkprobe --version or git commit)
  • OS and Rust toolchain if relevant
  • exact command line or config
  • steps to reproduce
  • impact (what an attacker could do)

You can also use GitHub Security Advisories for private disclosure.

We aim to acknowledge reports within a few days and will coordinate on a fix and disclosure timeline.

In scope

Issues we treat as security vulnerabilities include:

  • remote code execution or command injection in linkprobe itself (e.g. via crafted server responses, CLI arguments, or file paths)
  • memory-safety bugs in Rust code that are exploitable
  • authentication or authorization flaws in the --listen HTTP server beyond its intended read-only /metrics scrape
  • unintended credential disclosure (e.g. MQTT password logged or written to metrics output)

Out of scope

These are expected behavior, misconfiguration, or third-party issues — please use regular GitHub issues instead:

  • measuring against servers you do not trust (LibreSpeed hosts, custom --servers-url, iperf3 endpoints)
  • bugs or vulnerabilities in the external iperf3 binary
  • denial of service from running bandwidth tests (that is what the tool does)
  • binding --listen to a public interface without a firewall (bind 127.0.0.1 unless you intend to expose metrics)
  • missing TLS on MQTT when you point at a plain mqtt:// broker
  • feature requests and non-security bugs

Safe defaults

  • Prefer --listen 127.0.0.1:9090 unless Prometheus runs on the same host or network segment.
  • Treat --mqtt-password like any secret; avoid shell history and shared logs.
  • Only probe servers and broker URLs you control or explicitly trust.

There aren't any published security advisories