Skip to content

Security: Sebastianhayashi/chatgpt-web-ux-e2e-tester

Security

SECURITY.md

Security and Privacy

Treat project ZIPs as untrusted code

The Skill installs dependencies and starts user-provided applications in an isolated ChatGPT execution environment. A project ZIP can contain malicious install scripts, build scripts, test code, or runtime behavior. Review the project source and use the Skill only in an environment intended for untrusted execution.

Never include

  • production credentials or API keys;
  • real customer or regulated data;
  • payment credentials;
  • live OAuth client secrets;
  • production database dumps;
  • private certificates or signing material.

Use test-only fixtures, disposable accounts, local emulators, and .env.example values.

Authentication and external controls

The Skill may create or seed a local test account through repository-provided fixtures, a test API, or a disposable local database. It must not bypass live CAPTCHA, OTP, OAuth, payment, anti-abuse, or third-party security controls.

Report disclosure risk

Reports may include source excerpts, local paths, request metadata, visible test data, screenshots, and patches. Review and sanitize all artifacts before sharing publicly.

Reporting a vulnerability

Do not open a public issue containing exploit details, credentials, or private project code. Contact the repository owner through a private GitHub security advisory when available.

There aren't any published security advisories