Skip to content

Security: sigrix-io/mullion

Security

SECURITY.md

Security Policy

Mullion decodes untrusted bytes. That is the whole of its attack surface and it is worth naming plainly: every caller hands it a file somebody else chose.

Reporting a vulnerability

Do not open a public issue.

Email security@sigrix.io with enough detail to reproduce, and attach the input if you can. You will get an acknowledgement within 72 hours and an assessment as soon as we have one. CONTRIBUTING.md publishes the queues everything else waits in; security reports jump all of them.

What is in scope

  • An input that makes this library do something a caller cannot defend against — unbounded memory or time from a small file, a path that escapes the caller's control, an exception type that is not what the documented contract implies.
  • Anything that makes the safe spelling less safe than the unsafe one. The argument of this library is that open_bytes(data) is a better answer than Image.open(BytesIO(data)).convert("RGB"). A way in which it is worse is the most valuable report we can receive.
  • Leaking through the output. Metadata that should not survive normalization, or a code path that returns a caller another caller's pixels.
  • Defects in the repository itself: the CI workflows, the packaging, the published wheel's contents.

What is not in scope

  • Vulnerabilities in Pillow. Report those to python-pillow/Pillow. If a Pillow weakness is reachable in a way this library makes worse — a default we chose, a guard we removed — that is in scope here; say so.
  • Decompression bombs, in the general case. Pillow's own Image.MAX_IMAGE_PIXELS guard is active by default and this library does not raise or disable it. A path here that bypasses it is a finding.
  • Resource limits a caller must impose. How large an upload to accept, and how long to allow, are the application's decisions and cannot be made from inside a decoding library.

Supported versions

Version Status
0.1.x Pre-release — supported

Mullion is pre-1.0, and VERSIONING.md is blunt about what that means: nothing is stable, and any 0.x release may break any other. There is no back-porting, because there is nothing to back-port to. A finding lands in the next release with a CHANGELOG.md entry recording it.

What happens next

We will agree a disclosure timeline with you rather than impose one, and the changelog entry will credit you unless you would rather it did not.

This is a small team, and CONTRIBUTING.md is candid about what that means everywhere else. Security is the one queue where the 72-hour acknowledgement is a commitment rather than an aspiration.

There aren't any published security advisories