Security Policy
Supported Versions
| Version | Supported |
|---|---|
| 0.1.x | Yes |
Reporting a Vulnerability
If you discover a security vulnerability, please report it responsibly:
- Do not open a public GitHub issue
- Email the maintainers or use GitHub's private vulnerability reporting
- Include a description of the vulnerability, steps to reproduce, and potential impact
We will acknowledge your report within 48 hours and aim to provide a fix within 7 days for critical issues.
Scope
This policy covers:
- The Neural MRI Scanner backend (FastAPI server)
- The frontend application
- Docker deployment configurations
- CI/CD pipeline configurations
Out of Scope
- Vulnerabilities in upstream dependencies (report these to the respective projects)
- The HuggingFace Spaces demo deployment (best-effort only)
- Model weights and their outputs (these are third-party artifacts)
Security Considerations
Neural MRI Scanner runs model inference locally. Keep in mind:
- HuggingFace tokens: Store tokens in
.envfiles (gitignored) or environment variables. Never commit tokens. - Model loading: Only load models from trusted sources. The HF Hub search feature loads arbitrary model weights.
- Network exposure: The default configuration binds to localhost. If exposing to a network, use a reverse proxy with authentication.