Skip to content

Security: lizhiyao/oh-my-knowledge

Security

SECURITY.md

Security Policy

This policy is maintained in English only: vulnerability reports come from anywhere, and GitHub surfaces exactly one SECURITY.md with no locale variant.

Supported versions

The 1.x line is the only supported line: it is what receives security fixes, and a fix ships in the next release from main. 0.x is no longer supported and is not patched, including for security reports.

While the 1.x line is still prerelease, the supported build is the newest 1.0.0-beta.*, which npm serves from the next dist-tag — @latest still resolves to the last 0.x release:

npm i -g oh-my-knowledge@next

No manual tag move is needed: the release workflow derives the dist-tag from the version it publishes, so the first stable 1.0.0 release lands on latest and oh-my-knowledge@latest then installs a supported version. Report vulnerabilities against the build you run, including its npm view oh-my-knowledge dist-tags value when the channel matters.

Reporting a vulnerability

Do not open a public GitHub issue for security reports.

Use one of the following private channels:

  1. GitHub Security Advisories (preferred) — Report a vulnerability
  2. Email — coderdancestudio@gmail.com

Please include:

  • A description of the issue and its impact
  • Steps to reproduce (or a proof-of-concept)
  • The version of oh-my-knowledge affected
  • Your suggested fix, if any

Response expectations

This is a single-maintainer open-source project with no SLA. Reports are handled promptly on a best-effort basis:

  • Acknowledgement: typically within 3–5 days
  • Initial assessment: typically within 2 weeks
  • Fix timeline: depends on severity and complexity; critical issues prioritized

If you don't receive a response within 2 weeks, feel free to re-send or ping via GitHub.

Scope

This policy covers code in this repository (the oh-my-knowledge CLI and library). It does NOT cover:

  • Third-party executors you configure (Claude / OpenAI / Gemini CLIs or SDKs)
  • Custom assertion .mjs files you author (these execute in your Node process — treat them as trusted code you wrote)
  • MCP servers you configure in .mcp.json

The tool is designed for local trusted environments (dev machines, CI). See the "Security notice" section in README.md for the threat model.

Disclosure

Once a fix is shipped, the reporter is credited in the release notes unless they request anonymity.

There aren't any published security advisories