Skip to content

Security: KatherLab/MediSwarm

SECURITY.md

Security Policy

MediSwarm is a swarm/federated learning framework used by the ODELIA and DECADE medical-imaging consortia. Because it is deployed across hospital sites, we take security reports seriously.

Reporting a vulnerability

Please do not open a public issue for security problems.

Report privately through GitHub's "Report a vulnerability" button on the Security tab (Security → Advisories → Report a vulnerability). This opens a private advisory visible only to you and the maintainers.

Please include:

  • a description of the issue and its impact,
  • steps to reproduce (a minimal proof-of-concept if possible),
  • affected version / commit and pipeline (ODELIA or STAMP), and
  • any suggested remediation.

We aim to acknowledge a report within 5 working days and to agree a disclosure timeline with you. Please give us a reasonable window to release a fix before any public disclosure.

Scope

In scope:

  • the framework code in this repository (application/, scripts/, docker_config/, provisioning, and the startup-kit tooling),
  • the container images built from this repository, and
  • the CI/CD workflows.

Out of scope:

  • vulnerabilities in upstream dependencies (report those upstream — e.g. NVFlare), though we are glad to be told,
  • issues that require an already-compromised site host or valid VPN credentials, and
  • site-specific data, credentials, or network configuration (those live outside this repository and are managed by each participating institution).

MediSwarm is designed so that raw patient data never leaves a site — only model updates are exchanged. Reports demonstrating a way to exfiltrate local data or training inputs through the framework are especially welcome.

Supported versions

Security fixes target the latest released version on main. Older releases are not maintained; please upgrade to the current release before reporting.

There aren't any published security advisories