Skip to content

Security: Andy20010101/GPT-CodexPROJ

Security

SECURITY.md

Security Policy

Supported Surface

Security fixes should target the latest main branch state unless maintainers explicitly announce another supported release line.

Sensitive Areas

Be careful with:

  • logged-in browser sessions used by chatgpt-web-bridge
  • exported review artifacts and debug snapshots
  • local environment files and bridge/browser endpoints
  • self-improvement and recovery scripts that operate on live artifacts

Do not publish cookies, tokens, screenshots containing private data, or full local artifact bundles in public issues.

Reporting

If you discover a vulnerability:

  1. Avoid opening a public issue with exploit details.
  2. Prefer a private GitHub security advisory or another private maintainer contact path if one is available.
  3. Include a minimal reproduction, affected files or routes, impact, and any required local preconditions.
  4. Scrub credentials, session identifiers, and private artifact contents before sharing logs.

Scope Limits

This repository intentionally does not support:

  • credential bypass
  • CAPTCHA bypass
  • browser-login automation as a security workaround
  • access-circumvention features

Reports that depend on those behaviors should be framed as unsupported behavior, not accepted feature requests.

There aren't any published security advisories