Red team: November 2026

未关闭
#1,494 0 条评论 0 个 reaction 已指派 0 人 在 GitHub 查看

维护者通常 1 天内回复

还没有人认领这个 Issue。

评估

难度
5/5
预计耗时
一周以上
新手友好度
5/100
Issue 类型
文档
描述清晰度
需要澄清
活跃度
活跃
领域
security

调研方向

This is a maintainer tracking issue for a monthly red-team pass, not a code change. The process is defined in docs/specs/security-audit.md under "Red team"; read that section first. Done means every checklist item is ticked, confirmed findings are fixed or documented, and the summary sections below are filled in. Only maintainers with access to the private ledger can do this work.

由索引模型根据 Issue 内容生成。

描述

red-team

The monthly red-team pass (docs/specs/security-audit.md → "Red team"): an open-ended review of the code and the published guarantees, without the FAIL IF lines, on at least two model families. Findings go to the private ledger first; this issue records the pass once its findings are fixed or documented and disclosed.

Checklist
  • Build one stripped tree per engine from main, and record the commit reviewed.
  • Run each engine: Claude through Anthropic's Cyber Verification Program, Codex through OpenAI's Daybreak program (optionally Gemini).
  • Run each engine's second round, which proves its claims against the real code.
  • Verify every claimed break with a separate reviewer.
  • File each confirmed item in the ledger, fix it in a plain-worded PR, and close it citing the PR.
  • Record the pass in the private red-team/runs/.
  • After the nightly audit discloses the fixed findings, summarize the pass here and close.
The most serious findings

Nothing found allowed an account takeover or remote access to a terminal. Confirm or replace this sentence, and say why it holds. Everything rated above low is described below. Which release fixes it, and which releases were affected.

Severity What could happen Who could do it Fixed in
Severity What could happen, in one bolded sentence. How, in plain words. Who could do it. PR

Low-severity findings, fixed or documented: each in a clause with its PR, or none.

This month
  • Passes: each pass's commit and engines.
  • Fixed: PRs.
  • Documented as not defended or a known gap: PRs.
  • Disclosed: security-audit-finding.
  • Process changes it led to: PRs, or none.
主要语言
TypeScript
星标
5
派生
1
平均合并
13 小时 37 分钟
30 天内合并 PR
450

环境准备

这个项目没有提供开发容器、Dockerfile 或贡献指南,环境需要你自己搭建:先看它的 README,通用步骤见我们的新手贡献指南。

从这里开始

  1. 先读完整个 Issue,再读项目的贡献指南。
  2. 在 Issue 下留言说明你要接手 —— 这能避免两个人做同样的事。
  3. Fork 仓库,在一个分支上完成修改。
  4. 提交 Pull Request,并在描述里引用这个 Issue 编号。

diffplug/dormouse 的其他 Issue

查看 diffplug/dormouse 的全部 Issue

相似的 Issue

更多 TypeScript Issue

把新 issue 发到你的邮箱

精选适合新手参与的 GitHub issue 摘要。