Skip to content

Security: Cassandra-Collective/CASSANDRA

SECURITY.md

Security policy

This policy covers every repository in the Cassandra Collective organisation that does not have a SECURITY.md of its own. A repository's own policy takes precedence: Ephesus has one.

Reporting a vulnerability

Do not open a public issue, pull request or comment for a security problem.

Report it privately, in either of two ways.

  1. GitHub's private vulnerability reporting (preferred). It is enabled on every repository in the organisation. Open the repository's Security tab and select Report a vulnerability, or go straight to the form:

    Repository Private report
    CASSANDRA Report a vulnerability
    CASSANDRA-Chorus Report a vulnerability
    Dervish Report a vulnerability
    Ephesus Report a vulnerability
    .github (profile and brand files) Report a vulnerability

    For any other repository, the form is at https://github.com/Cassandra-Collective/<repository>/security/advisories/new. You need a GitHub account to use it.

  2. Email to sensoymertefe@gmail.com, with SECURITY in the subject line.

What to include

  • The repository, and the commit, branch or release you looked at.
  • What the problem is and where it is: file, function, workflow or configuration.
  • Steps to reproduce it, or a proof of concept.
  • What you think the impact is: who could do what, and under which conditions.
  • Whether anyone else knows about it, and whether you plan to publish it, and when.

Describe secrets and personal data rather than pasting them. If you found a live token or key, say where it is; do not use it.

What to expect

The Cassandra Collective is a small organisation run by volunteers. There is no security team and no guaranteed response time. What we aim for:

  • A person reads every report and replies. We aim to acknowledge a report within a few days; it can take longer.
  • We tell you whether we could reproduce the problem, and keep you informed while it is fixed.
  • When a fix is ready, we publish a GitHub security advisory where it is warranted, crediting you unless you would rather stay anonymous.

Scope

In scope: the code, configuration, workflows and published files of the organisation's repositories. For example:

  • a secret, token or key committed to a repository or its history;
  • code that loads, imports or executes an untrusted file or network input in an unsafe way;
  • a GitHub Actions workflow that an outside pull request or issue can abuse;
  • a dependency vulnerability that our code actually makes reachable.

Out of scope:

  • Vulnerabilities in third-party software or services themselves. Report those to their maintainers, and tell us if our use of them makes the problem worse.
  • What a model trained by these projects writes (wrong, biased or unsafe text). That is a research finding or a bug: open an ordinary issue. It is a security problem only if it exposes private data or lets someone run code or reach systems they should not.
  • Problems that need an attacker who already controls your machine or your account.

Supported versions

The projects are research and pre-release software. Security fixes go to each repository's default branch (main) only, unless that repository's own policy says otherwise.

Disclosure

Please give us a reasonable time to ship a fix before you publish anything. We coordinate disclosure through GitHub security advisories.

There aren't any published security advisories