Skip to content

Security: Microrobotics-Simulation-Framework/MADDENING

Security

SECURITY.md

Security Policy

Reporting a Vulnerability

If you discover a security vulnerability in MADDENING, please report it responsibly:

Email: nick@microrobotica.org

Please include:

  • Description of the vulnerability
  • Steps to reproduce
  • Affected versions
  • Any suggested fix (optional)

Response Timeline

  • Acknowledgement: within 3 business days
  • Initial assessment: within 7 business days
  • Fix or mitigation: best effort, prioritised by severity

Supported Versions

Version Supported
0.1.x Yes

Dependency Monitoring

MADDENING monitors its core dependencies for known security vulnerabilities using two mechanisms:

  1. GitHub Dependabot — enabled on the repository; automatically monitors PyPI dependencies for published CVEs and creates pull requests for security-relevant version bumps.

  2. Manual changelog review — JAX ecosystem libraries (JAX, jaxlib, Equinox, Optax) are reviewed at each MADDENING release for correctness-affecting changes (XLA compiler changes, numerical behaviour changes) that may not be classified as security vulnerabilities.

Security-relevant dependency updates are flagged in the Security section of CHANGELOG.md.

Scope

MADDENING is a computation library, not a network service. The primary security concerns are:

  • Supply chain: malicious or compromised dependencies
  • Numerical correctness: silent computation errors (see known_anomalies.yaml)
  • Denial of service: pathological inputs that cause excessive memory or compute usage

MADDENING assumes trusted inputs (Section 2 of docs/regulatory/intended_use.md). Input sanitization and validation is the responsibility of the downstream integration layer.

There aren't any published security advisories