Skip to content

Security: headmesh-dev/headmesh

Security

.github/SECURITY.md

Security policy

Headmesh controls who can reach which machines, so isolation bugs are treated as critical.

Reporting a vulnerability

Please use GitHub's private vulnerability reporting on headmesh-dev/headmesh ("Security" → "Report a vulnerability"), or email security@headmesh.dev. Do not open public issues for security problems.

We aim to acknowledge reports within 2 business days and to ship a fix or mitigation for critical issues (cross-tenant reachability, authentication bypass, key disclosure) within 7 days.

Scope

In scope: the control server, cell agent, CLI, sandbox daemon and image, plugins, and the hosted service at *.headmesh.dev. Out of scope: vulnerabilities in Headscale or Tailscale themselves (report those upstream; we will help coordinate), and denial of service through volumetric traffic.

Guarantees we test for

  • A device in one organisation can never reach a device in another unless an active, time-limited grant allows exactly that device and port.
  • Revoking a grant or removing a device takes effect within seconds and removes keys and routes.
  • Every enrollment, grant and revocation produces a tamper-evident audit entry.

There aren't any published security advisories