Skip to content

Security: BlackBearCC/libu-studio

Security

SECURITY.md

Security Policy

Supported versions

libu-studio is pre-1.0; only main is supported. Pin to a commit if you need stability.

Reporting a vulnerability

Please do not open a public GitHub issue for security-sensitive reports.

Use one of:

  1. GitHub Security Advisories (preferred): https://github.com/BlackBearCC/libu-studio/security/advisories/new
  2. Email the maintainer (see GitHub profile)

We will acknowledge within 7 days and aim to publish a fix or mitigation within 30 days for confirmed vulnerabilities.

What counts as a vulnerability here

This is a content-generation toolchain, not a server. The realistic threat model is small. Reportable issues include:

  • Credential leaks in shipped code or lab.dump.sql examples — anything resembling gho_*, ghs_*, pat_*, Bearer <token>, liblib session cookies, browser local storage dumps
  • Arbitrary file overwrite — a script that writes outside its declared output directory based on attacker-controlled input (manifest path, slug, filename)
  • Command injection — anywhere a shell command interpolates a value from the manifest or DB without quoting
  • Path traversal in lab.db lookups (e.g. a ..-bearing inject_target_dir causing writes outside the target project)
  • Insecure download — fetching from a URL stored in the DB without validating the host

Out of scope

  • The repo intentionally has no auth surface and runs entirely local
  • liblib.art account abuse / token theft is a liblib.art concern, not ours
  • macOS sandbox bypasses in bgrm.swift (Apple Vision is sandboxed by the OS)
  • Bugs in upstream Python deps (Pillow, numpy, imageio_ffmpeg) — please report those upstream

Safe disclosure

If you confirm a vulnerability and want it cited, we will credit you in the fix commit and the published advisory. Coordinated disclosure with downstream forks: we will give you the embargo window you request, up to 90 days.

There aren't any published security advisories