Skip to content

Security: FlowLabsWorks/NetherNetProxy

SECURITY.md

Security policy

Reporting a vulnerability

Please report security problems privately through GitHub: Security → Report a vulnerability on the repository page. Do not open a public issue for them. You will get an answer as soon as possible; this is a small hobby project maintained in spare time.

Only the latest release is supported.

Trust model

What the proxy does and does not protect:

  • It never sees, stores or forwards the player's Microsoft identity. The Login packet is ignored, and the real server authenticates the player after the Transfer packet.
  • The LAN side is unauthenticated by design: any device in the LAN can ask the proxy for a Transfer to the configured server. That is the same exposure as the server's own address. Run the proxy only in networks where this is acceptable.
  • LAN discovery uses a fixed, publicly known key (the "application ID" of the vanilla game). It provides no secrecy.
  • Each connection is protected by DTLS with a throw-away self-signed certificate, checked against the fingerprint in the console's offer.
  • The service runs as an unprivileged user with a hardened systemd unit (no capabilities, read-only filesystem, only IPv4/IPv6 sockets). Concurrent connection setups are capped and message sizes are bounded.

Releases

Release archives and SHA256SUMS are built by the release workflow from a tagged commit and carry a signed build provenance attestation (Sigstore). Verify a download with gh attestation verify <file> --repo FlowLabsWorks/NetherNetProxy. The repository is configured so that only the maintainer (and later explicitly invited collaborators) can push; everyone else contributes through pull requests.

In scope

Memory safety or denial-of-service problems reachable from the LAN (malformed discovery, STUN, DTLS/SCTP or Bedrock packets), and mistakes in the unit or install script that grant more access than described above.

Out of scope

Vulnerabilities in Minecraft, in Bedrock Dedicated Server, or in your own network configuration, and issues that need an attacker who already controls the proxy host.

There aren't any published security advisories