Skip to content

Security: brandonhon/scry

Security

SECURITY.md

Security policy

Supported versions

main and the latest tagged release receive security fixes.

Reporting a vulnerability

Please don't open a public GitHub issue for security reports.

Instead, open a private security advisory so we can triage and patch before details are public.

Include:

  • the version / commit you tested
  • a minimal reproduction
  • the impact (what an attacker can do)
  • any suggested remediation

You'll get a first acknowledgement within 72 hours. High-severity issues will ship a patch release as soon as a fix is verified; medium/low issues fold into the next scheduled release.

Scope

scry is an active-probe network tool. By design it opens TCP connections, sends packets, and can run user-supplied Lua scripts. If an attacker can pass flags to a running scry process or feed it a malicious script, they can already do what they want on the network scry has access to. The security boundary is the scry process itself, not the network it probes.

In scope:

  • Memory-unsafe patterns in the Go source
  • Lua sandbox escapes that let a script read/write arbitrary files, run processes, or touch sockets outside the scanner's own handles
  • Config-file parsing vulnerabilities
  • CLI argument parsing vulnerabilities
  • Dependency CVEs that affect scry's running binary

Out of scope:

  • "scry can scan networks" — that is the feature, not a vulnerability.
  • User running scry against a host without authorisation — that's on the operator.
  • Hardcoded values (like the default progress bar throttle) — they're intentional.

Fixed advisories

None yet. v0.1.0 is the first tagged release.

There aren’t any published security advisories