The org's standard furniture for a new repo: licence, code of conduct, security policy, contributing guide, issue and pull-request templates, and the ignore rules. No module yet — that arrives as the first pull request, so this branch exists to open one against. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_016wDDVXWMDz82WqE1i969r4
3.6 KiB
Security Policy
Thank you for helping keep Runic Gateway and its users safe.
Reporting a vulnerability
Please do not report security vulnerabilities through public issues, pull requests, or the wiki. A public report tips off attackers before a fix is available.
Instead, report privately by email to:
Please include as much of the following as you can:
- The repository and component affected.
- The type of issue (e.g. authentication bypass, injection, secret exposure, remote code execution, denial of service).
- Step-by-step instructions to reproduce, and a proof-of-concept if you have one.
- The impact — what an attacker could do with it.
- Any suggested remediation.
You will receive an acknowledgement of your report, typically within a few days. We will keep you informed as we investigate and work toward a fix, and we are happy to credit you in the release notes once the issue is resolved (let us know if you would prefer to remain anonymous).
Scope
Runic Gateway is a self-hosted platform made up of several components:
| Component | Repo | Network exposure |
|---|---|---|
| Website core (site + admin + API) | RunicGateway/website |
Internet-facing (behind a reverse proxy) |
| Module-Rust (this repo) | RunicGateway/Module-Rust |
No listener of its own — runs inside the website process |
| rust-link sidecar | RunicGateway/Rust-Link |
The only network-facing part of the game bridge |
| Oxide bridge plugin | RunicGateway/Rust-Plugins |
Loopback only — dials the sidecar on 127.0.0.1 |
| Installer | RunicGateway/installer |
Not a service — an operator-run deployment tool |
| Documentation | RunicGateway/docs |
Content only |
Because instances are self-hosted, the security of any given deployment also
depends on how it is configured and operated — strong secrets (JWT_SECRET,
SECRET_ENC_KEY, database and admin passwords), a correctly configured reverse
proxy and TRUST_PROXY, and keeping the game server itself unreachable from the
internet (only the sidecar should be exposed). See each repo's README for the
security model.
Module-specific notes
Three things follow from what a module is, and they are policy rather than oversight:
- The module boundary is not a security boundary. A module runs in the same Node process as core with the same privileges, and its schema fragment runs against the same database. It is a code-organisation and distribution boundary. Installing a module is the same trust decision as installing the site itself — which is appropriate for a self-hosted operator choosing their own software, and is why module installation is admin-only. "A module can reach core internals" is therefore not a vulnerability report; "an unprivileged user can install or enable a module" very much is.
- Access control lives in core, not in the module and never in the sidecar.
Route protection is core's
requireAuth/requireRolemiddleware, and what a visitor is allowed to see of live game state is this module's own visibility layer. A module route that reaches game data without going through those is a security bug worth reporting. - This module handles every server's sidecar auth token. Each is encrypted at rest with
core's
secretBoxand is write-only in the API — never returned to any client, in any shape. Anything that would echo it back, log it, or expose it to the browser is a security issue.
Supported versions
This project is developed continuously and does not maintain long-term release
branches. Security fixes land on main; please run a recent build.