Skip to content

Security: lvis-project/lvis-app

Security

SECURITY.md

Security Policy

Supported versions

Only the latest released version on the GitHub Releases page receives security fixes. There is no LTS branch.

Reporting a vulnerability

Please report security vulnerabilities privately — do not open a public GitHub issue.

Contact: jo.dreame@gmail.com

Include:

  • LVIS version (or git commit hash if running from source)
  • Affected component (host app, plugin, marketplace, MCP, etc.)
  • Reproduction steps + proof-of-concept if possible
  • Impact assessment (information disclosure, code execution, privilege escalation, denial of service)
  • Your preferred disclosure timeline

What to expect

Phase Target
Acknowledgement within 5 business days
Initial triage + severity within 10 business days
Fix released or mitigation documented depends on severity (see below)

Severity targets:

  • Critical (remote code execution, sandbox escape, credential theft): patch within 14 days
  • High (privilege escalation, sensitive data exposure): patch within 30 days
  • Medium / Low: addressed in the next regular release cycle

Scope

In scope:

  • LVIS desktop app (this repository)
  • The plugin SDK and bundled plugins shipped with each release
  • Marketplace fetcher behavior and bundled plugin verification
  • Tool execution sandbox, permission gates, and approval flow
  • Auto-update verification (signature, manifest, integrity)

Out of scope (handle directly via upstream):

  • Electron framework vulnerabilities → Electron security team
  • Node.js / V8 / Chromium runtime issues → respective upstreams
  • Operating system or third-party software interactions
  • Vulnerabilities in third-party plugins not bundled with the release

Coordinated disclosure

We follow a 90-day disclosure deadline by default, adjustable by mutual agreement based on patch availability and risk to users. We credit reporters in release notes unless anonymity is requested.

Notes on the security model

  • All on-disk data under ~/.lvis/ uses 0o700 for directories and 0o600 for files. Secrets are additionally encrypted via the OS keychain (safeStorage).
  • Tool execution passes through a multi-stage permission gate (approval-gate, permission-manager, sandbox runner on macOS/Windows where available).
  • Marketplace artifacts are verified against signed manifests when the corp-portal-template capability is configured; verify the threat model in docs/architecture/permission-policy-design.md before deploying in a regulated environment.

If you find a defect in any of these guarantees, please report it.

There aren't any published security advisories