Skip to content

Security: FLIMKit/.github

Security

SECURITY.md

Security Policy

This is the default policy for repositories in the FLIMKit organisation that do not carry their own. FLIMKit itself has a fuller one, and it takes precedence for that repository.

These are academic research tools maintained by Alex Hunt at the Centre for Inflammation Research, University of Edinburgh, not commercial products with a security team behind them. Reports are handled on a best-effort basis, but they are taken seriously and they will get a reply.

Reporting a vulnerability

Do not open a public issue for a security problem.

Use one of these instead:

  1. GitHub's private vulnerability reporting, from the Security tab of the repository in question.
  2. Email alexander.hunt@ed.ac.uk with FLIMKit security in the subject line.

Please include what the problem is and what an attacker gets out of it, the version and platform you saw it on, and steps to reproduce. If a malformed input file triggers it, a minimal file or a generator script is far more useful than a description.

Expect acknowledgement within 7 days and an assessment within 30. This is a single-maintainer project, so if you have not heard back in that window, send a follow-up rather than assuming it was ignored. Credit goes in the advisory unless you would rather stay anonymous.

Scope

These tools run locally on data you point them at. The realistic risk is a hostile or corrupt input file, not a remote attacker.

In scope: code execution, memory corruption or path traversal from parsing a data file; code execution or file overwrite from loading a session, project or config file; insecure handling of credentials or signing material in build scripts.

Out of scope: crashes or unhandled exceptions on malformed files with no path to code execution or data loss, which are ordinary bugs and still worth an issue; running out of memory on a large file; anything requiring an attacker who already has your user account on the machine.

A plugin loaded from ~/.flimkit/plugins is ordinary Python running with your privileges. There is no sandbox, and a plugin misbehaving is not a vulnerability. A way to get one loaded without the setting being on is.

There aren't any published security advisories