Skip to content

Security: devforai-creator/Safe-UGC-UI

SECURITY.md

Security Policy

Scope

This project focuses on safely rendering untrusted UGC (user‑generated UI). The security surface in scope includes:

  • Card validation (packages/validator)
  • Rendering safeguards (packages/react)
  • Type/schema definitions that drive validation (packages/types, packages/schema)

Out of scope (but still important for adopters):

  • Host application logic and data sources
  • Asset hosting/CDN configuration and network policies
  • Prompting/LLM output quality and post‑processing
  • Third‑party dependency vulnerabilities unless they directly impact the library’s security guarantees

Threat Model (High‑Level)

  • Goal: prevent untrusted UI from escaping its container, loading unsafe external content, or abusing layout to spoof/overlay host UI.
  • Non‑goals: full sandboxing of browser execution, general-purpose runtime expression evaluation, or protection against compromised host apps.
  • Status: best‑effort; not formally audited.

Reporting Vulnerabilities

Please avoid public disclosure until a fix is available.

Preferred:

  • If hosted on GitHub, use Security Advisories to report privately.

Fallback:

  • Open an issue labeled security with minimal details and request a private channel.

If no private channel is available, contact the maintainer listed in README.md (if present).

Response Expectations

  • We aim to acknowledge reports within 7 days and provide a fix or mitigation plan as soon as practical.
  • No formal SLA is currently offered.

Safe Usage Guidelines (Consumers)

  • Always validate cards before rendering.
  • Use the provided asset allowlist (@assets/) and reject external URLs.
  • Keep renderer/validator versions in sync.
  • Treat JSON Schema as structural validation only; do not rely on it for security.

There aren't any published security advisories