Skip to content

Security: denialwm/denial

SECURITY.md

Security policy

Denial is a compositor and therefore sits on security-sensitive boundaries: it handles input, application surfaces, clipboard data, screen capture, session state, and signed software distribution.

Supported versions

Version Security support
Latest tagged release Supported
main Fixes are developed here, but it is not a released package
Older tagged releases Not supported

During the public beta, users should update to the latest tagged release to receive security fixes.

Reporting a vulnerability

Report suspected vulnerabilities through GitHub private vulnerability reporting. Do not publish vulnerability details in an issue, discussion, pull request, or other public channel before a fix and disclosure are coordinated.

Please include as much of the following as possible:

  • the affected Denial version, commit, and installation method;
  • the operating system, kernel, GPU, driver, and relevant hardware;
  • a description of the security impact and the boundary that was crossed;
  • minimal steps or a proof of concept that reproduces the behavior;
  • logs, traces, or screenshots with unrelated private information removed;
  • whether the issue is known to be actively exploited; and
  • any disclosure deadline or prior disclosure that the maintainer should know about.

Do not include credentials, signing material, personal data, or unrelated secrets.

What to report

Security reports may include:

  • lock-screen, authentication, or session-boundary bypasses;
  • unintended input observation or injection;
  • unauthorized access to application surfaces, clipboard contents, screen capture, or other session data;
  • memory-safety problems with a security impact;
  • ways for an untrusted Wayland or Xwayland client to gain capabilities beyond the intended protocol boundary;
  • arbitrary code execution or privilege escalation caused by Denial;
  • package, update, repository, signing, or release-integrity failures; and
  • denial-of-service behavior that can be triggered intentionally across a security boundary.

Ordinary crashes, visual defects, compatibility problems, and performance regressions without a security impact should be reported through the normal issue tracker.

Security boundaries

  • User-provided Flutter shell bundles are trusted local code. They are not currently sandboxed from the compositor or the user's session.
  • Live UI development exposes an authenticated Dart VM service only on loopback and stores its discovery URI in a user-private runtime file. The packaged editor profile is non-pausing, but another process already running as the same user remains inside the trusted-user boundary and may use the VM service directly.
  • A system already controlled by an administrator, root process, compromised kernel, or compromised user account is outside Denial's threat boundary.
  • A vulnerability entirely within an upstream project, with no Denial-specific behavior or impact, should be reported to that upstream project. If it is exploitable through Denial, it may also be reported here so the project can coordinate mitigation and dependency updates.
  • Unsupported hardware or configurations are not automatically security issues, but crossing a security boundary remains reportable regardless of support status.

Response and disclosure

Doctor Logix will make a reasonable effort to:

  1. acknowledge a complete private report within seven days;
  2. reproduce and assess its impact;
  3. coordinate a fix, release, and disclosure with the reporter; and
  4. credit the reporter if they want public credit.

This is a volunteer-maintained project and the acknowledgement target is not a contractual service-level guarantee. Complex fixes, upstream coordination, or hardware-specific investigation may take longer.

Please allow a reasonable remediation period before disclosure. If the report is accepted, coordination will continue through the private advisory until a release is available or the parties agree that disclosure is appropriate.

Safe research

Only test systems and data you own or are explicitly authorized to use. Avoid privacy violations, data destruction, persistence, service disruption beyond what is necessary to demonstrate the issue, and access to data unrelated to the report.

Denial does not currently operate a bug bounty program and cannot promise payment or other compensation for reports.

There aren't any published security advisories