Skip to content

Security: khakhasshi/OptionWorkstation

SECURITY.md

Security Policy

Supported Versions

Option Workstation is pre-1.0. Security fixes target the latest main branch and the most recent tagged release only.

Version Supported
latest release yes
main best effort
older releases no

Reporting a Vulnerability

Do not open a public issue for a vulnerability.

After the repository is published, use GitHub Private Vulnerability Reporting from the repository Security tab. Before that feature is enabled, email jiangjingzhe2004@gmail.com with the subject Option Workstation Security.

Include:

  • affected commit or release;
  • component and route;
  • reproducible steps using fake credentials and a paper account only;
  • impact and prerequisites;
  • any suggested mitigation.

Do not include a real token, account number, licensed market-data sample, or order history. If a credential was exposed while investigating, revoke it before sending the report.

We aim to acknowledge a complete report within seven days and provide an initial assessment within fourteen days. Timelines may change for a maintainer-run project, but reporters will receive status updates when a private contact channel is available.

Security Boundaries

  • The default server binding is 127.0.0.1.
  • The application has no built-in multi-user authentication or TLS.
  • Longbridge credentials are expected to remain in process memory only.
  • Real-money order submission is intentionally unsupported.
  • Paper-order submission is server-disabled by default and separately gated.
  • Historical market data and audit ledgers are local operator assets.

Running the server on a shared host, reverse proxy, or public network without additional authentication and isolation is outside the supported threat model. See docs/THREAT_MODEL.md.

Out of Scope

The following are not security vulnerabilities by themselves:

  • market losses, missed fills, slippage, or inaccurate investment decisions;
  • stale or unavailable third-party market data that is visibly reported;
  • model approximation error described in docs/MODEL_LIMITATIONS.md;
  • attacks requiring prior control of the local operating-system account;
  • provider outages, entitlement limits, or rate limiting.

Data-quality failures that are hidden or mislabeled are still important; report those with the data-quality issue form.

There aren't any published security advisories