Skip to content

Security: golovanov-dev/weft

Security

SECURITY.md

Security Policy

Weft sits in front of every request a site serves and handles its forms. A vulnerability in it is a vulnerability in every site built on it — so please report one, and please report it privately first.

Supported versions

Version Supported
0.2.x ✅ current

Before 1.0.0 only the latest minor line is supported. A fix ships as a patch release of that line and is not backported: upgrading within 0.x is the fix.

Reporting a vulnerability

Do not open a public issue. Use GitHub's Report a vulnerability button on this repository. It opens a private advisory that only the maintainers can see, and the fix and the published advisory come out of the same thread.

Please include:

  • what you did, in enough detail to reproduce it;
  • the Weft version (composer show golovanov/weft) and PHP version;
  • what an attacker gains;
  • anything you already know about a fix.

You will get an acknowledgement within 3 working days and an assessment within 10. If a report turns out to be a real vulnerability, we will agree a disclosure date with you, credit you in the advisory and the changelog unless you would rather stay anonymous, and ship the fix in a patch release of the supported line.

If you do not hear back within the acknowledgement window, please chase it — a missed email is far more likely than a decision to ignore you.

Scope

In scope: the engine in src/ — routing and path handling, request parsing, CSRF, rate limiting, captcha verification, security headers, JWT verification, request tracing and its redaction, the database wrapper, and the alert client.

Out of scope, because they are the site's job:

  • Code a site writes on top of Weft: its pages, API endpoints and services.
  • A site's deployment: the web server, TLS, the document root, file permissions and secrets in config/config.php.
  • The example/ site as a deployed application — it is documentation, not a hardened product. Engine defects it demonstrates are in scope.
  • Findings from an automated scanner with no demonstrated impact.

There aren't any published security advisories