Skip to content

Security: killerboyyy777/Machine-Learning-mmo

Security

SECURITY.md

Security Policy

Supported Versions

Version Supported
master ✅
tags v* (latest) ✅
older tags ❌ (upgrade first, then report)

Only the current master branch and the latest tagged release receive security fixes.

Reporting a Vulnerability

Do not open a public issue for security problems. Instead:

  1. Go to the repository's Security tab > Report a vulnerability (private advisory), or
  2. Contact the maintainer directly via the email on their GitHub profile.

Include:

  • What you did, step by step (commands sent, rooms/NPCs involved)
  • What you expected vs what happened (privilege escalation, data loss, remote crash, treasury/market theft, auth-token bypass, ...)
  • server.py revision (git rev-parse HEAD), config files in use (server_config.json, ml_config.json), relevant logs
  • Whether it needs other players/bots online to trigger

You will get an initial response within 7 days. If the issue is confirmed, we will coordinate a fix and credit you in the release notes (unless you prefer to stay anonymous).

Disclosure Policy

  • We ask for up to 90 days of coordinated disclosure before you publish details of a confirmed vulnerability.
  • Once a fix is released, details may be published freely.
  • Cheat/exploit reports that only affect game balance (not security -- e.g. a profitable market loop) belong in public issues with repro steps, not here. When in doubt, report privately and we will route it.

Scope Notes

  • The game server trusts its local operator completely: scores.json, world.json, and the config files are operator-controlled. A malicious operator is out of scope.
  • Denial-of-service via sheer connection volume is rate-limited by TEXTMMO_MAX_CONNECTIONS, not eliminated; report bypasses of it here.
  • The dashboard HTTP server (start_dashboard) is loopback-oriented tooling, not hardened public infrastructure.

There aren't any published security advisories