Skip to content

Security: sidkandan/GeminiDisplayKit

Security

SECURITY.md

Security Policy

Supported versions

This is a hackathon project, currently at 0.1.0. While we aim to keep this published code safe to run, do not deploy this framework to production environments holding real user data or API keys without your own security review.

Version Supported
0.1.x ✓ (best-effort)

Reporting a vulnerability

If you discover a security vulnerability, please do not open a public GitHub issue. Instead:

  1. Email siddharth.kandan@gmail.com with a description of the issue
  2. Allow up to 7 days for an initial response
  3. We will work with you on a fix and a coordinated disclosure timeline

What counts as a vulnerability

  • API key leakage paths (e.g., a code path that prints GEMINI_API_KEY to logs)
  • SSRF / path-traversal in the bridge's static file serving
  • Cross-site scripting in display HUDs (these load user-controlled Gemini output)
  • Code execution in the agent-sandbox path that escapes Google's sandbox
  • Anything that lets a remote caller dump server-side state

What does NOT count

  • "The Cloudflare quick-tunnel is public" — by design; that's the install path
  • "You can POST any image to /api/analyze-media" — by design; demos require it
  • "The bridge accepts unauthenticated requests" — by design for the demo model; document an auth pattern in your own deployment

Acknowledgements

We'll credit reporters in SECURITY-HALL-OF-FAME.md (created when the first report comes in) unless you prefer to remain anonymous.

There aren't any published security advisories