Skip to content

Security: john-lawniczak/OpenSourceOrtho

Security

SECURITY.md

Security Policy

Scope and safety boundary

OpenSource Ortho is a clear-aligner planning safety playground and research toolkit, not a medical device and not complete treatment-planning software. It does not diagnose, prescribe, approve treatment, decide treatment safety or suitability, authorize physical use, or replace a licensed dental professional. Any use, manufacturing, or physical application of outputs is the user's own responsibility and risk. Please read docs/SAFETY.md. Reports that a finding "should have approved/cleared a plan" are out of scope by design - the software intentionally never makes those claims.

In-scope security issues include, for example:

  • ways to make model-generated output bypass the lint_finding() safety gate
  • path traversal, SSRF, or unsafe file handling in orthoplan/server.py or the mesh/print/export paths
  • leakage of an AI-connector API key (it must stay session-only and never be persisted, logged, or echoed back)
  • any path that stores or transmits patient-derived data contrary to the stated boundary

Reporting a vulnerability

Please do not open a public issue for a security vulnerability.

  • Preferred: open a private security advisory from the repository's Security tab.
  • If private advisories are unavailable, contact the active project maintainer through their current public project profile with steps to reproduce and impact.

You can expect an acknowledgement within a few days. This is a volunteer-run open-source project, so timelines are best-effort. Please give us a reasonable window to fix an issue before public disclosure.

Supported versions

Only the latest release on main is supported. There are no backported fixes.

There aren't any published security advisories