Skip to content

Security: acunli/ClipContext

Security

SECURITY.md

Security Policy

Reporting a vulnerability

If you find a security issue in ClipContext — an auth bypass, a way to access another user's data, a credential leak, an injection vector, or similar — please do not open a public GitHub issue.

Instead, use GitHub's private vulnerability reporting for this repository (Security tab → Report a vulnerability), or email the maintainer directly if that isn't available. Include:

  • What you found and where (file/endpoint if known).
  • Steps to reproduce.
  • What you think the impact is.

Expect an acknowledgement within a few days. This is a hackathon-origin project maintained on a best-effort basis — there's no formal SLA, but security reports get priority over feature work.

Scope

In scope: the FastAPI backend (src/), the Next.js frontend (frontend/), and how they integrate with Firebase, YouTube OAuth, Fireworks, and the AMD vLLM inference path.

Out of scope: vulnerabilities in third-party services themselves (Firebase, Google OAuth, Fireworks, YouTube Data API) — report those to the relevant vendor.

Known, accepted limitations

These are documented tradeoffs, not vulnerabilities — see docs/Backend.md and README.md "Known limitations" for the full list. The short version: job state and YouTube session/token storage are in-memory (single-instance only, by design, not an oversight), and this is a hackathon-scope project without a security audit — treat a public deployment accordingly (see docs/Deployment.md for hardening notes).

Handling secrets

Never commit .env, frontend/.env.local, or any Firebase service-account JSON file. See docs/Environment.md for what every credential does and where it belongs. If you accidentally commit a secret, rotate it immediately at the provider (Fireworks, Google Cloud, Firebase) — removing it from a later commit does not remove it from git history.

There aren't any published security advisories