zetetic-team-subagents

遵循以下最佳实践的项目将能够自愿的自我认证,并显示他们已经实现了核心基础设施计划(OpenSSF)徽章。

没有一套可以保证软件永远不会有缺陷或漏洞的做法;如果规范或假设是错误的,即使合适的方法也可能失败。也没有哪些做法可以保证一个项目能够维持健康和运作良好的开发者社区。但是,遵循最佳做法可以帮助改善项目的成果。例如,一些做法可以在发布之前进行多人评估,这可以帮助您找到其他难以找到的技术漏洞,并帮助建立信任,并希望不同公司的开发人员之间进行重复的交互。要获得徽章,必须满足所有“必须”和“禁止”的条款,满足所有“应该”条款或有合适的理由,所有“建议”条款必须满足或未满足(至少希望考虑)。欢迎通过 GitHub网站创建问题或提出请求进行反馈。另外还有一个一般讨论邮件列表

如果这是您的项目,请在您的项目页面上显示您的徽章状态!徽章状态如下所示: 项目13847的徽章级别为passing 这里是如何嵌入它:
您可以通过将其嵌入在您的Markdown文件中:
[![OpenSSF Best Practices](https://www.bestpractices.dev/projects/13847/badge)](https://www.bestpractices.dev/projects/13847)
或将其嵌入到HTML中来显示您的徽章状态:
<a href="https://www.bestpractices.dev/projects/13847"><img src="https://www.bestpractices.dev/projects/13847/badge"></a>


这些是白银级别条款。您还可以查看通过黄金级别条款。

Baseline Series: 基准等级1 基准等级2 基准等级3

        

 基本 12/17

  • 常规

    请注意,其他项目可能使用相同的名称。

    11 problem-shaped skills backed by 97 sourced reasoning patterns — Curie to Toulmin, as Claude Code agents. Every claim cites its source; a pre-commit gate blocks unsourced constants. Install the gates alone in 30s (zetetic-gates). The only agent system where "I don't know" is a feature.

    请使用 SPDX许可证表达格式;例子包括“Apache-2.0”,“BSD-2-Clause”,“BSD-3-Clause”,“GPL-2.0+”,“LGPL-3.0 +”,“MIT”和“(BSD-2-Clause OR Ruby)”。
    如果有多种语言,请将它们列为逗号分隔值(可选空格),并将它们从最多到最少使用。如果有长列表,请至少列出前三个最常见的列表。如果没有语言(例如,这是仅文档或仅测试项目),请使用单个字符“ - ”。请使用每种语言的常规大小写,例如“JavaScript”。
    通用平台枚举(CPE)是用于信息技术系统,软件和软件包的结构化命名方案。在报告漏洞时,它可用于多个系统和数据库。
  • 先决条件


    项目必须拥有通过徽章。 [achieve_passing]

  • 基本项目网站内容


    关于如何贡献的信息必须包括对可接受的贡献的要求(例如,引用任何所需的编码标准)。 (需要网址) [contribution_requirements]

    CONTRIBUTING.md states the requirements a contribution must meet: the coding standard it is held to (https://github.com/cdeust/zetetic-team-subagents/blob/main/rules/coding-standards.md), the five sections every new agent definition must answer, the hook rules, and the testing expectation. https://github.com/cdeust/zetetic-team-subagents/blob/main/CONTRIBUTING.md


  • 项目监督


    该项目应该有一个合法机制,所有对项目软件做出显著贡献的开发人员都声明他们有合法授权作出这些贡献。最常用且易于实施的方法是使用开发者原产地证书(DCO),用户可以在其提交中添加“signed-off-by”,另外,项目链接到DCO网站。本条款也可以使用贡献者许可协议(CLA)或其他法律机制实现。 (需要网址) [dco]
    DCO是推荐的机制,因为它易于实现,在源代码中进行跟踪,并且git使用“commit -s”直接支持“签名”功能。更有效的是,项目文件解释了该项目的“签名”手段。 CLA是一项法律协议,用于定义知识产权许可给组织或项目的条款。捐助者转让协议(CAA)是将知识产权权利转让给另一方的法律协议;项目不必具备CAA,因为CAA增加了潜在贡献者不愿贡献的风险,特别是如果接收者是一个营利性组织。 Apache Software Foundation CLA(个人贡献者许可证和公司CLA)是CLA的示例,用于确定项目的这些CLA的风险对项目的影响小于其获益。

    There is no DCO or CLA requirement today. All non-trivial code is written by the sole maintainer, who owns the copyright and licenses it under MIT; the authorship statement is at https://github.com/cdeust/zetetic-team-subagents#license. A DCO sign-off requirement is the intended mechanism once outside contributors start submitting non-trivial changes.



    该项目必须明确定义和记录其项目治理模式(决策方式,包括关键角色)。 (需要网址) [governance]
    需要有一些成熟的书面方式来作出决定和解决争端。在小项目中,这可能就像“项目拥有者和负责人做所有最终决定”一样简单。有各种治理模式,包括仁慈的独裁者和正式的精英统治;有关详细信息,请参阅治理模式。集中式(例如,单一维护者)和分散式(例如,组维护者)方法都已经在项目中成功使用。治理信息不需要记录创建项目分支的可能性,因为对于FLOSS项目来说总是可能的。

    There is no GOVERNANCE.md. CONTRIBUTING.md documents how to submit a change and CODE_OF_CONDUCT.md documents how to behave, but neither states who decides, what a change needs to land, or how disagreements resolve. Tracked with acceptance criteria in issue #68.



    该项目必须采取行为守则,并将其发布在标准的位置。 (需要网址) [code_of_conduct]
    项目可能能够提高社区的文明程度,并通过采取行为守则来规定对可接受的行为的期望。这可以帮助在发生问题之前避免问题,并使项目成为更加欢迎的地方来鼓励贡献。这应该只关注项目社区/工作场所的行为。行为规范的示例是贡献者约定行为准则和Linux内核代码冲突

    https://github.com/cdeust/zetetic-team-subagents/blob/main/CODE_OF_CONDUCT.md, at the standard repository location, with reporting and enforcement sections. The reporting sentence points at a package.json that does not exist in this repository; that broken pointer is tracked in issue #73, and the private channel in https://github.com/cdeust/zetetic-team-subagents/blob/main/SECURITY.md works today.



    该项目必须明确定义和公开记录项目中的关键角色及其职责,包括这些角色必须执行的任务。必须清楚指出谁是什么角色,尽管这可能没有以相同的方式记录下来。 (需要网址) [roles_responsibilities]
    治理和角色和职责的文档可以在一个地方。

    The key roles and who holds them are not publicly documented. In practice there is one maintainer who reviews, merges and releases, but the criterion asks for that to be written down, and it is not. Tracked in issue #68.



    如果任何一个开发者丧失工作能力或丧生,该项目必须能够继续保持最小的中断。特别是,在确认个人无行为能力或死亡的一周内,项目必须能够创建和关闭问题,接受提议的更改和发布软件版本。这可以通过确保其他人有任何必要的密钥,密码和合法权利来继续该项目。运行FLOSS项目的个人可以通过在锁箱中提供密钥,并提供任何所需的合法权利(例如DNS名称)来实现。 (需要网址) [access_continuity]

    Not documented. Some continuity properties hold by construction (the source is public under MIT, every release is tagged and Sigstore-verifiable, so anyone can fork and continue), but issue creation, change acceptance and release publication all depend on one account today and no succession arrangement is recorded. Tracked in issue #68.



    项目必须具有2个或更多的“公交车因子”。 (需要网址) [bus_factor]
    “公交车系数”(又名“卡车因子”)是指最少数量的项目成员,如果突然离开项目(“被公交车撞了”),项目会由于缺乏具备知识的或有能力的人员而暂停。 卡车因子 工具可以对GitHub上的项目进行估计。有关详细信息,请参阅Cosentino等人的评估Git存储库的卡车因子

    The bus factor is 1. This is a single-maintainer project, stated plainly rather than papered over. The same constraint is already recorded in https://github.com/cdeust/zetetic-team-subagents/blob/main/docs/SCORECARD.md, which explains why the Scorecard Code-Review and Branch-Protection checks are structurally unreachable here: GitHub does not permit self-approval, so requiring reviews would deadlock every merge. Review rigor is carried instead by the always-on hard CI gates.


  • 文档


    该项目必须有一个文档化的路线图,描述项目打算做什么,至少在下一年做什么。 (需要网址) [documentation_roadmap]
    项目可能没有实现路线图,没关系。路线图的目的是帮助潜在的用户和贡献者了解项目的预期方向。它不需要特别详细。

    There is no roadmap document. The README's "What this system does not do" section states current limits, which is not a forward plan, and the CHANGELOG is backward-looking by construction. Tracked with acceptance criteria in issue #69.



    该项目必须包括项目生成的软件的架构(也称为高级别设计)的文档。如果项目不产生软件,请选择“不适用”(N/A)。 (需要网址) [documentation_architecture]
    软件架构解释了程序的基本结构,即程序的主要组件,它们之间的关系以及这些组件和关系的关键属性。

    The architecture is documented at two levels. The component model and how the pieces compose is in the README (https://github.com/cdeust/zetetic-team-subagents#what-you-actually-get and https://github.com/cdeust/zetetic-team-subagents#knowledge-ingestion--a-query-indexed-semantic-layer, which gives the web_ingest to manifest_gate to semantic_layer pipeline and states that the three tools import none of each other, with the skill as the single wiring point). The agent file shape, frontmatter contract and routing mechanism are specified in https://github.com/cdeust/zetetic-team-subagents/blob/main/docs/AGENT-INTERNALS.md, and the memory subsystem's scopes, ACL and MCP surface in https://github.com/cdeust/zetetic-team-subagents/blob/main/docs/MEMORY-MCP.md



    该项目必须书面记录用户从项目生成的软件的安全性上可以获得和不能指望的内容(其“安全需求”)。 (需要网址) [documentation_security]
    这些是软件意图满足的安全需求。

    https://github.com/cdeust/zetetic-team-subagents/blob/main/SECURITY.md states what the user can and cannot expect. It opens by naming what installing the plugin grants (session-execution rights, no sandbox between a modified file and the machine), lists the supply-chain assurances and how to verify each, and closes with an explicit "What this does NOT claim" paragraph: provenance proves who built the bundle and from which commit, not that the scripts are free of defects, and it is worth nothing to a user who never runs the verification.



    该项目必须为新用户提供“快速启动”指南,以帮助他们快速使用该软件。 (需要网址) [documentation_quick_start]
    这个想法是向用户展示如何入门,使软件完全可以做事情。这对于潜在用户是至关重要的。

    Two commands and the system is installed: https://github.com/cdeust/zetetic-team-subagents#install. A narrower quick start for the gates alone (claude plugin install zetetic-gates) is on the same page, and the first thing to do with it is the reproducible one-liner at https://github.com/cdeust/zetetic-team-subagents#the-system-enforcing-its-own-standard



    项目必须努力使文档与当前版本的项目结果(包括项目生成的软件)保持一致。任何已知的文档缺陷使其不一致必须修正。如果文档基本是最新的,但是错误地包括一些不再是真实的旧信息,那么将其视为缺陷,然后像往常一样进行跟踪和修复。 [documentation_current]
    该文档可以包括有关软件版本之间的差异或更改的信息,与/或链接到旧版本的文档。这个条款的意图在于努力保持文档的一致性,而不是文档必须完美。

    Known inconsistencies exist, so the honest answer is unmet until they are fixed. Measured on 2026-07-28 the tree holds 97 genius agents plus 23 team agents, 76 skill files and 19 hook registrations, while README claims 119 agents in its badge and body and 118 in its footer, CONTRIBUTING claims 22 team agents, 64 skills and 18 hooks, and .claude-plugin/marketplace.json claims 78 skills and 20 hooks. Separately, all five test commands in CONTRIBUTING.md name files that do not exist, and CODE_OF_CONDUCT.md routes conduct reports to a package.json this repository does not have. Tracked in issue #72 (counts, with a claim gate so it stops recurring) and issue #73 (broken pointers).



    项目存储代码库首页和/或网站必须在获得成就的48小时内标识并超链接任何成就,包括此最佳实践徽章。 (需要网址) [documentation_achievements]
    一个成就是项目致力于满足的任何外部条款,包括徽章。该信息不需要在项目网站首页上。使用GitHub的项目可以通过将其添加到README文件中,将成果放在存储代码库首页。

    The repository front page carries a badge row that identifies and hyperlinks the project's achievements: the live OpenSSF Best Practices badge for this project (https://www.bestpractices.dev/projects/13847), live CI status, the test, agent, skill and hook counts, and the MIT license (https://github.com/cdeust/zetetic-team-subagents#readme). The badge is dynamic, so the level it displays tracks this questionnaire without a further edit. Per-release achievements are recorded in https://github.com/cdeust/zetetic-team-subagents/blob/main/CHANGELOG.md and the security posture in https://github.com/cdeust/zetetic-team-subagents/blob/main/docs/SCORECARD.md


  • 无障碍和国际化


    该项目(项目网站和项目成果)都应遵循无障碍最佳做法,使残疾人仍然可以参与项目,并在合理的情况下使用项目成果。 [accessibility_best_practices]
    对于Web应用程序,请参阅 Web 内容无障碍指导 (WCAG 2.0,英文) 以及其支持文档 理解 WCAG 2.0(英文);或者 W3C 无障碍信息(英文). 图形界面(GUI)应用考虑使用环境特定的无障碍指导(例如 Gnome, KDE, XFCE, Android, iOS, Mac, 以及 Windows). 一些TUI应用 (例如,`ncurses` 程序) 可以做一些特定的工作,增强可访问性(例如,`alpine` 的 `force-arrow-cursor` 设置)。多数命令行应用没有特别的无障碍设置。本条款经常是不涉及(N/A),例如,对于组件库。以下是一些要采取的行动或需要考虑的问题的例子:
    • 提供非文字内容的文字替代,以便于转换为其他需要的格式,如大字体、盲文、语音、符号或者简单文字( WCAG 2.0 指导 1.1 (英文))
    • 颜色不被用作传达信息,指示动作,提示响应或区分视觉元素的唯一视觉方式。 ( WCAG 2.0 指导 1.4.1(英文))
    • 文本和文字图像的视觉呈现对比度至少为4.5:1,除了大文本,次要文本和标识符之外。 ( WCAG 2.0 指导 1.4.3(英文))
    • 使用键盘可访问所有功能 (WCAG 指导 2.1)
    • 图形界面应用(GUI)或者Web应用在目标平台上,应该至少使用一种屏幕阅读器测试(例如,NVDA;Jaws;Windows上的WindowEyes;Mac & iOS上的VoiceOver;Linux/BSD上的Orca;Android上的TalkBack)。TUI程序可以减少过度渲染以防止屏幕阅读器的冗余读取。

    The project results have no graphical interface: they are Markdown definitions and command-line scripts whose output is plain text rendered by the host terminal. The project site is GitHub and the documentation is Markdown; the one image on the front page, the banner, carries a text alternative (alt="Zetetic Agents: 97 reasoning patterns, one epistemic standard").



    该项目产生的软件应该被国际化,以便为目标受众的文化,地区或语言进行简单的本地化。如果国际化(i18n)不适用(例如,该软件不会生成针对最终用户的文本,并且不排序可读文本),请选择“不适用”(N/A)。 [internationalization]
    本地化“是指适应产品,应用或文档内容以满足特定目标市场(语言环境)的语言,文化和其他要求。国际化是“设计和开发产品,应用或文档内容,使不同文化,地区或语言的目标受众更容易本地化”。 (请参阅 W3C的“本地化与国际化”)。软件只需通过国际化即可达到此标准。不需要另一种特定语言的本地化,因为一旦软件已被国际化,其他人就可以从事本地化。

    The tools' user-facing strings (checker findings, hook notices, setup output) are English literals and are not externalized for localization, and the agent and skill definitions are English prose. Externalizing them has not been done, so this is reported unmet rather than claimed.


  • 其他


    如果项目站点(网站,存储库和下载URL)存储用于外部用户认证的密码,则必须使用密钥拉伸(迭代)算法将密码存储为具有每用户盐值的迭代散列(例如,PBKDF2,Bcrypt或Scrypt)。如果项目站点不存储密码,请选择“不适用”(N/A)。 [sites_password_security]
    请注意,使用 GitHub 符合此条款。此条款仅适用于用于外部用户认证到项目站点的密码。如果项目站点必须登录到其他站点,则可能需要为此目的存储密码(因为使用像Bcrypt这样的算法会使这些密码无效)。本条款将 crypto_password_storage 条款应用于项目站点,类似于sites_https条款。

    The project operates no site of its own that authenticates external users. The repository and the download URLs are GitHub, and no password is stored by this project.


 变更控制 1/1

 报告 3/3

  • 错误报告流程


    项目必须使用问题跟踪器来跟踪每个问题。 [report_tracker]
  • 漏洞报告流程


    除了要求匿名的报告者外,该项目必须对过去12个月内解决的所有漏洞报告的报告者表示感谢。如果过去12个月没有修复漏洞,请选择“不适用”(N/A)。 (需要网址) [vulnerability_report_credit]

    No vulnerability has been resolved in the last 12 months, so there is no reporter to credit. The commitment is documented: reporters are credited in the release notes for the patched version unless they request anonymity. https://github.com/cdeust/zetetic-team-subagents/blob/main/SECURITY.md



    该项目必须有一个书面的流程来响应漏洞报告。 (需要网址) [vulnerability_response_process]
    这与security_report_process有很强的相关性,它需要有一个书面的流程来报告漏洞。它还涉及到spam_report_response,它需要在一定时间内响应漏洞报告。

    https://github.com/cdeust/zetetic-team-subagents/blob/main/SECURITY.md documents the whole response process: the private advisory channel, what a report must contain, a per-severity SLA table (24 hours first response and 7 days to patch for critical, 3 and 14 days for high), a five-step disclosure timeline with a 30-day default coordinated date, and what is out of scope.


 质量 18/19

  • 编码标准


    该项目必须确定其使用的主要语言的具体编码风格指南,并要求贡献一般情况下符合此要求。 (需要网址) [coding_standards]
    在大多数情况下,这是通过参考一些现有的风格指南来完成的,可能列出差异。这些风格指南可以包括提高可读性的方法和减少缺陷可能性(包括漏洞)的方法。许多编程语言有一个或多个广泛使用的风格指南。样式指南的示例包括 Google风格指南(英文) SEI CERT编码标准(英文)

    The standard is a document in the tree, not a preference: https://github.com/cdeust/zetetic-team-subagents/blob/main/rules/coding-standards.md covers SOLID, the layer dependency rule, size limits, dependency injection, local reasoning, source discipline, mutation testing, the definition of done and the boy-scout rule, each with its primary source. CONTRIBUTING.md names it as binding and excerpts the load-bearing rules: https://github.com/cdeust/zetetic-team-subagents/blob/main/CONTRIBUTING.md



    如果至少有一个FLOSS工具可以应用于所选择的语言,项目必须自动执行其选定的编码风格。 [coding_standards_enforced]
    这可以使用静态分析工具和/或通过代码重新格式化强制代码实现。在许多情况下,工具配置包含在项目的存储库中(因为不同的项目可能会选择不同的配置)。项目可以允许风格例外(通常会);在发生例外的情况下(应该很少),它们必须在其位置的代码中记录在案,以便可以对这些例外进行检视,并使工具能够在将来自动处理它们。这些工具的例子包括ESLint(JavaScript)和Rubocop(Ruby)。

    Enforced automatically at two points. At commit time https://github.com/cdeust/zetetic-team-subagents/blob/main/hooks/pre-commit-zetetic.sh runs tools/zetetic-checker.sh on the staged diff and returns exit 2 on a violation. In CI four hard gates block the merge (https://github.com/cdeust/zetetic-team-subagents/blob/main/.github/workflows/ci.yml): the craftsmanship checker on newly added files, the redaction sweep on the full tree under ZETETIC_PROFILE=strict, the tools regression suite, and shellcheck at error severity. The ratchet shape (new files gated, legacy debt reported) is documented in the workflow itself.


  • 可工作的构建系统


    本地二进制文件的构建系统必须遵守传递给它们的相关编译器和链接器(环境)变量(例如CC,CFLAGS,CXX,CXXFLAGS和LDFLAGS),并将它们传递给编译器和链接器。构建系统可以使用附加标志来扩展它们,但不能简单地用自己的替换提供的值。如果没有生成本地二进制文件,请选择“不适用”(N/A)。 [build_standard_variables]
    应该很容易启用特殊的构建功能,如地址消毒剂(Address Sanitizer,ASAN),或符合发布加固最佳实践(例如,通过轻松打开编译器标志来实现)。

    No native binaries are generated. There is no compiler or linker invocation anywhere in the project, so CC, CFLAGS and LDFLAGS have nothing to apply to.



    构建和安装系统在在相关标志中要求时,应该保留调试信息(例如,“install -s”未被使用)。如果没有构建或安装系统(例如典型的JavaScript库),请选择“不适用”(N/A)。 [build_preserve_debug]
    例如,如果使用这些语言,则应该设置CFLAGS(C)或CXXFLAGS(C ++)创建相关的调试信息,并且在安装过程中不应该剥离它们。支持和分析时,需要调试信息,也可用于测量编译二进制文件中加固特性的存在。

    There is no build or installation system that strips anything. The software is Markdown and scripts, installed by copying, and the source on disk is what runs.



    如果子目录中存在交叉依赖关系,则由项目生成的软件的构建系统必须不能递归地构建子目录。如果没有构建或安装系统(例如典型的JavaScript库),请选择“不适用”(N/A)。 [build_non_recursive]
    项目构建系统的内部依赖关系信息需要准确,否则,对项目的更改可能无法正确构建。不正确的构建可能会导致缺陷(包括漏洞)。大型构建系统中的常见错误是使用“递归构建”或“递归生成”,即包含源文件的子目录的层次结构,其中每个子目录都是独立构建的。除非每个子目录完全独立,否则这是一个错误,因为依赖关系信息不正确。

    There is no build system, so there are no subdirectory builds to make recursive or otherwise.



    该项目必须能够重复从源代码文件生成的过程,并获得完全相同的比特位结果。如果没有发生构建(例如,直接使用源代码而不是编译使用的脚本语言),请选择“不适用”(N/A)。 [build_repeatable]
    GCC和clang用户可能会发现-frandom-seed选项有用;在某些情况下,这可以通过强制某种排序来解决。更多建议可以在可重复构建(英文)站点找到。

    Although nothing is compiled, the one thing that is generated is bit-for-bit reproducible by construction. https://github.com/cdeust/zetetic-team-subagents/blob/main/tools/build-release-bundle.sh creates the release tarball with --sort=name --mtime=@0 --owner=0 --group=0 --numeric-owner, so the archive does not depend on filesystem order, timestamps or the building user. The script says why: an attestation over a non-deterministic archive proves nothing a second builder can check.


  • 安装系统


    该项目必须提供一种使用常用惯例轻松安装和卸载由项目生成的软件的方法。 [installation_common]
    示例包括使用软件包管理器(在系统或语言级别),“make install/uninstall”(支持DESTDIR),标准格式的容器或标准格式的虚拟机映像。安装和卸载过程(例如,打包)可以由第三方FLOSS软件实现。

    Installed and uninstalled through the ecosystem's package manager: "claude plugin marketplace add cdeust/zetetic-team-subagents" then "claude plugin install zetetic-team-subagents" (https://github.com/cdeust/zetetic-team-subagents#install). The manual path is https://github.com/cdeust/zetetic-team-subagents/blob/main/scripts/setup.sh with install, update, uninstall and configure verbs, and the installer writes a manifest so that uninstall removes exactly what it added.



    最终用户的安装系统必须遵守用于在安装时选择构建工件写入位置的标准约定。例如,如果在POSIX系统上安装文件,它必须遵守DESTDIR环境变量。如果没有安装系统或没有标准惯例,请选择“不适用”(N/A)。 [installation_standard_variables]

    There is no installation system that writes built artifacts to a configurable prefix. Installation copies agents, skills, hooks and tools into the host's own plugin directory under ~/.claude, which is the convention the host defines, so DESTDIR and similar variables have nothing to select.



    该项目必须为潜在开发人员提供一种快速安装所有项目成果和支持环境所必需的环境,包括测试套件和测试环境。必须通过常规惯例执行。 [installation_development_quick]
    可以使用生成的容器和/或安装脚本来实现。外部依赖关系一般通过调用系统和/或语言包管理器(根据 external_dependencies 条款)进行安装。

    git clone and the software is complete: nothing is compiled and there are no dependencies to resolve. https://github.com/cdeust/zetetic-team-subagents/blob/main/CONTRIBUTING.md lists the prerequisites (bash 4+, GNU coreutils, ripgrep, python3) and https://github.com/cdeust/zetetic-team-subagents/blob/main/.github/workflows/ci.yml gives the exact command for every suite, which is the same command a developer runs locally. The stale command list inside CONTRIBUTING is being corrected under issue #73.


  • 外部维护的组件


    项目必须以计算机可处理的方式列出外部依赖关系。 (需要网址) [external_dependencies]
    通常这是使用包管理器和/或构建系统的约定完成的。请注意,这有助于实施 installation_development_quick

    Listed in two computer-processable forms. Every release publishes a CycloneDX SBOM (zetetic-team-subagents.cdx.json) inventorying every bundled file with its hash, including the vendored third-party skill material, alongside an EXECUTABLE-MANIFEST.sha256 that enumerates every hook and tool that will run on the user's machine (https://github.com/cdeust/zetetic-team-subagents/blob/main/SECURITY.md). The only external supply-chain surface, the GitHub Actions, is declared in https://github.com/cdeust/zetetic-team-subagents/blob/main/.github/dependabot.yml and each action is pinned to a full commit SHA in the workflows.



    项目必须监视或定期检查其外部依赖(包括便利副本)以检测已知的漏洞,并修复可利用的漏洞或将其验证为不可利用的漏洞。 [dependency_monitoring]
    这可以使用源分析器/依赖项检查工具/软件组成分析工具(例如OWASP的Dependency-CheckSonatype的Nexus AuditorSynopsys的Black Duck软件组成分析Bundler-audit(针对Ruby))来完成。一些程序包管理器包括执行此操作的机制。如果无法利用组件的漏洞,这是可以接受的,但是这种分析是困难的,有时简单地更新或修复零件就更容易。

    https://github.com/cdeust/zetetic-team-subagents/blob/main/.github/dependabot.yml runs weekly against the github-actions ecosystem, which is this project's only external dependency surface: there is no runtime package to install, so there is no pip or npm manifest to watch, and the config states that reasoning rather than declaring an empty ecosystem. Dependabot updates the pinned SHA and rewrites the version comment in place, so pinning discipline is preserved automatically. OpenSSF Scorecard runs on a schedule on top of it (https://github.com/cdeust/zetetic-team-subagents/blob/main/.github/workflows/scorecard.yml), with each finding's disposition recorded in https://github.com/cdeust/zetetic-team-subagents/blob/main/docs/SCORECARD.md



    该项目必须满足下述情况之一:
    1. 可以轻松识别和更新重用的外部维护组件;
    2. 使用系统或编程语言提供的标准组件。
    这样,如果在重用的组件中发现了一个漏洞,将容易更新该组件。 [updateable_reused_components]
    符合这一条款的典型方法是使用系统和编程语言的包管理系统。许多FLOSS程序与“便利库”一起分发,这些库是标准库的本地副本(可能是分支)。一般没问题。但是,如果程序*必须*使用这些本地(分支)副本,则“标准”库的安全更新将使这些附加副本仍然易受攻击。这对于基于云的系统尤其是一个问题;如果云提供商更新他们的“标准”库,但程序不会使用它们,那么这些更新实际上不会有帮助。参见,例如,“Chromium:为什么它不在Fedora中作为适当的包”(Tom Callaway)

    Reused components are identifiable and updateable. The GitHub Actions are pinned by full commit SHA with a version comment and updated by Dependabot weekly. The Python code imports only the standard library, so there is no vendored runtime dependency to go stale. The one convenience copy, the vendored writing and redaction rule material, carries its provenance in the file and is inventoried in the per-release CycloneDX SBOM.



    该项目应避免使用已弃用或过时的功能和API,如果FLOSS替代品在其使用的技术集合(“技术堆栈”)中以及项目支持的大多数用户中可用(以便用户可以随时访问该替代品)。 [interfaces_current]

    The technology stack is current and nothing deprecated is relied on. The Python is standard library only (argparse, hashlib, json, ssl, subprocess, urllib, xml.etree, tempfile) with no deprecated or removed module, and CI pins Python 3.12. The shell targets bash 4+ with POSIX coreutils, and shellcheck at error severity would flag obsolete constructs on every push.


  • 自动测试套件


    必须将自动测试套件应用于至少一个分支的共享代码库的每次签入。该测试套件必须生成关于测试成功或失败的报告。 [automated_integration_testing]
    这个要求可以被视为test_continuous_integration的一个子集,但是仅仅是测试,而不需要持续集成。

    Every push to main and every pull request runs the full gate set and reports pass or fail per job: seven jobs in https://github.com/cdeust/zetetic-team-subagents/blob/main/.github/workflows/ci.yml (memory suites, agent auditor, worktree tests, zetetic checker, craftsmanship checker, redaction sweep, tools regression suite covering 16 suites), plus shellcheck and CodeQL in their own workflows. Four of them are hard gates that block the merge. The Python pytest suites are the one part not yet wired into CI, tracked in issue #71.



    该项目必须为过去六个月内修复的至少50%的错误,在自动化测试套件中添加回归测试。 [regression_tests_added50]

    Well above 50 percent, because the standard makes it mandatory: rules/coding-standards.md section 13.1 G2 requires every bug fixed in a pull request to carry a regression test that fails on the pre-fix code. Recent fixes follow it: issue #64 (redaction-checker missing untracked files) shipped with the test that reproduces the false local pass, and issue #60's CodeQL fix was made structural so the checker itself is the regression guard. https://github.com/cdeust/zetetic-team-subagents/blob/main/rules/coding-standards.md



    如果有至少一个FLOSS工具可以以所选语言度量此条款,该项目的FLOSS自动测试套件必须具有至少80%语句覆盖率。 [test_statement_coverage80]
    许多FLOSS工具可用于度量测试覆盖范围,包括gcov / lcov,Blanket.js,Istanbul和JCov。请注意,满足这个条款并不能保证测试套件是完备的,而不满足该条款则意味着测试套件很差。

    Measured, not estimated. coverage.py 7.13.5 with pytest 9.0.2 on 2026-07-28 over tools/, hooks/ and scripts/: 2090 statements, 1628 missed, 22 percent, with 334 tests passing. Restricting the source set to the shipped surface of hooks/ plus tools/ and dropping the one-shot scripts/patch_*.py migrations gives 30 percent, so the answer does not change under a friendlier scoping. Three gate cores are near-complete (acceptance_gate 98 percent, manifest_gate 99 percent, semantic_layer 99 percent) while six files are at zero, including hooks/pre-tool-secret-shield.py and tools/memory-mcp-server.py. N/A is not available here: coverage.py is a FLOSS tool that measures Python, which is one of this project's languages. The full per-file measurement and the acceptance criteria are in issue #71.


  • 新功能测试


    该项目必须具有正式的书面策略,一旦添加了主要的新功能,新功能的测试必须被添加到自动测试套件中。 [test_policy_mandated]

    A formal written policy exists. rules/coding-standards.md section 13.1 A1 to A3 requires happy paths, every enumerated edge case and every failure path (including the emission of the signal itself) to map to a test, and section 13.2 makes the mapping a required artifact of the pull request. Section 13.3 enumerates the refusals, including that classifying a coverage gap in new code as a follow-up is blocking. https://github.com/cdeust/zetetic-team-subagents/blob/main/rules/coding-standards.md



    该项目必须在其关于变更建议的书面指导中包括要为主要新功能添加测试的策略。 [tests_documented_added]
    但是,只要在实践中添加了测试,即使是非正式规则也是可以接受的。

    CONTRIBUTING.md states it for hooks ("Changes to that hook need a corresponding test in tests/") and rules/coding-standards.md section 13 states it generally, with the Completion Ledger as the mechanism that makes it checkable in review. https://github.com/cdeust/zetetic-team-subagents/blob/main/CONTRIBUTING.md and https://github.com/cdeust/zetetic-team-subagents/blob/main/rules/coding-standards.md


  • 警告标志


    在实际允许时,项目必须最大限度地严格修复项目生成的软件中的警告。 [warnings_strict]
    某些项目无法有效启用某些警告。需要证明的是,项目正在努力的启用警告标志,以便早期发现错误。

    Maximally strict for the languages in use. shellcheck runs at error severity over every shell file in hooks/ and tools/ with no "|| true" escape, so a single finding fails the job: see the step "ShellCheck (error severity - hard gate)" in https://github.com/cdeust/zetetic-team-subagents/blob/main/.github/workflows/shellcheck.yml. CodeQL runs the widest query pack, security-and-quality, rather than the default security-only pack. The redaction sweep runs under ZETETIC_PROFILE=strict, which promotes every finding to exit 1 (https://github.com/cdeust/zetetic-team-subagents/blob/main/.github/workflows/ci.yml). Warning severity for shellcheck is currently informational, with the backlog measured and scheduled in issue #74.


 安全 11/13

  • 安全开发知识


    该项目必须实施安全设计原则(来自“know_secure_design”)(如适用)。如果项目不生产软件,请选择“不适用”(N/A)。 [implement_secure_design]
    例如,项目结果应该具有故障安全默认值(默认情况下,访问决策应该拒绝,默认情况下项目的安装应该是安全的),也应该有完全的仲裁(每一个可能被限制的访问权限必须被检查,不可绕过)。请注意,在某些情况下,原则会发生冲突,在这种情况下必须做出选择(例如,许多机制使事情更复杂,违反“机制经济”/“保持最简化”的原则)。

    The principles are implemented, with one stated exception. Fail-safe defaults: the memory ACL denies any write outside a scope's owner list and tools/manifest_gate.py rejects an ungrounded source with exit 3. Separation of privilege: the release job holds only id-token and attestations write, everything else is contents: read. Open design: the enforcement surface and its documented limits are public (https://github.com/cdeust/zetetic-team-subagents#what-this-system-does-not-do). Economy of mechanism: web_ingest, manifest_gate and semantic_layer import none of each other and are wired by a skill, so each is independently testable. The exception is complete mediation at one tool boundary, described under input_validation below.


  • 使用基础的良好加密实践

    请注意,某些软件不需要使用加密机制。

    由项目产生的软件中的默认安全机制不得取决于具有已知严重弱点(例如,SHA-1密码散列算法或SSH中的CBC模式)的加密算法或模式。 [crypto_weaknesses]
    CERT:SSH CBC漏洞中讨论了SSH中CBC模式的问题。

    The only digest in the tree is SHA-256 and the only transport is the platform default TLS context. No SHA-1, no MD5, and no hand-selected cipher mode.



    该项目应该支持多种加密算法,如果一个被破解,用户可以快速切换。普通的对称密钥算法包括AES,Twofish和Serpent。通用密码散列算法的选择包括SHA-2(包括SHA-224,SHA-256,SHA-384和SHA-512)和SHA-3。 [crypto_algorithm_agility]

    The project selects no cryptographic algorithm of its own. Transport algorithms are negotiated by the platform TLS stack and release signing is delegated to Sigstore, so there is no project-level choice for a user to switch.



    该项目必须支持在与其他信息(如配置文件,数据库和日志)分离的文件中存储身份验证凭据(如密码和动态令牌)以及私有加密密钥,并允许用户更新和替换它们,而无需重新编译代码。如果项目从不处理身份验证凭据和私有加密密钥,请选择“不适用”(N/A)。 [crypto_credential_agility]

    The software processes no authentication credentials and holds no private keys. Release signing is keyless through Sigstore, where the signing certificate is short-lived and issued against the workflow's OIDC identity, so there is no stored private key to separate out or rotate.



    该项目产生的软件应该支持所有网络通信的安全协议,如SSHv2或更高版本,TLS1.2或更高版本(HTTPS),IPsec,SFTP和SNMPv3。默认情况下,FTP,HTTP,Telnet,SSLv3或更早版本和SSHv1等不安全协议将被禁用,只有在用户专门配置时才启用。如果项目生成的软件不支持网络通信,请选择“不适用”(N/A)。 [crypto_used_network]

    Outbound fetches use the platform TLS stack correctly, but insecure schemes are not disabled by default. tools/web_ingest.py hands a caller-supplied URL to the standard library client without restricting the scheme first, so plain HTTP is accepted on the same footing as HTTPS. The fix is an allowlist at that boundary, re-applied after each redirect hop, with plain HTTP available only behind an explicit opt-in. Reported unmet rather than argued around.



    项目生成的软件(如果支持或使用TLS)应该至少支持TLS版本1.2。请注意,TLS的前身称为SSL。如果软件不使用TLS,请选择“不适用”(N/A)。 [crypto_tls12]

    Outbound TLS goes through ssl.create_default_context (https://github.com/cdeust/zetetic-team-subagents/blob/main/tools/web_ingest.py), whose minimum protocol version on the Python versions this project targets is TLS 1.2. The context is created once and reused for every fetch, including the robots.txt fetch, so no request escapes it.



    由项目生成的软件必须,如果它支持TLS,则在使用TLS(包括子资源)时默认执行TLS证书验证。如果软件不使用TLS,请选择“不适用”(N / A)。 [crypto_certificate_verification]

    Certificate verification is on by default and is never disabled. ssl.create_default_context sets verify_mode to CERT_REQUIRED and check_hostname to True, and there is no CERT_NONE, no check_hostname = False and no unverified context anywhere in the tree. The context is preferentially built against certifi's CA bundle when available, falling back to the system store. https://github.com/cdeust/zetetic-team-subagents/blob/main/tools/web_ingest.py



    项目生成的软件(如果支持TLS)必须在发送具有私有信息(如安全Cookie)的HTTP头之前执行证书验证。如果软件不使用TLS,请选择“不适用”(N/A)。 [crypto_verification_private]

    No private information is ever sent. The outbound client transmits only a User-Agent, Accept-Encoding and the conditional-GET validators (If-None-Match, If-Modified-Since); it sends no cookie, no authorization header and no credential. Certificate verification is performed by the default context before any request body or header leaves, and no fetch is made outside that context.


  • 安全发布


    该项目必须加密签名旨在广泛使用的项目结果的发布,并且必须有一个书面流程,向用户解释如何获取公共签名密钥并验证签名。这些签名的私钥不得在项目网站上直接向公众发布。如果发行版本不适用于广泛使用,请选择“不适用”(N/A)。 [signed_releases]
    项目结果包括源代码和适用的任何生成的可交付成果(例如,可执行文件,包和容器)。生成的交付项可以单独签名源代码。可以用签名的git标签实现(使用加密数字签名)。项目可以从git类似的工具分别提供生成的结果,但在这些情况下,单独的结果必须单独签署。

    Every release is cryptographically signed through Sigstore and the verification path is documented. The release workflow attests the tarball, the executable manifest and the CycloneDX SBOM with actions/attest-build-provenance under a least-privilege OIDC identity, and self-verifies with the shipped verifier before publishing (https://github.com/cdeust/zetetic-team-subagents/blob/main/.github/workflows/release.yml). Users verify with "gh attestation verify zetetic-team-subagents.tar.gz --repo cdeust/zetetic-team-subagents", documented at https://github.com/cdeust/zetetic-team-subagents/blob/main/SECURITY.md. Signing is keyless, so no private key exists to be stored on the distribution site: the certificate is short-lived, issued by Fulcio against the workflow's identity, and logged in the Rekor transparency log.



    建议在版本控制系统中,每个重要版本标签(作为主要版本的一部分的标签,次要版本或修复公开提到的漏洞)都按照signed_releases中的要求进行加密签名,并可验证。 [version_tags_signed]

    Release tags are annotated but not cryptographically signed: "git tag -v v2.35.0" reports no signature. The release artifacts are covered instead, each with a SHA-256 companion and a Sigstore build-provenance attestation (https://github.com/cdeust/zetetic-team-subagents/blob/main/SECURITY.md). Signing the tags themselves would add a second, independent anchor and has not been set up.


  • 其他安全问题


    项目结果必须检查来自潜在不受信任来源的所有输入,以确保它们有效( *白名单*),如果对数据有限制,则拒绝无效输入。 [input_validation]
    请注意,将输入与“不良格式”(*黑名单*)的列表进行比较通常是不够的,因为攻击者通常可以绕过黑名单。例如,数字被转换成内部格式,然后检查它们是否在最小和最大(包括)之间,并且检查文本字符串以确保它们是有效的文本模式(例如,有效的UTF-8,长度,语法等)。一些数据可能需要是“任何东西”(例如,文件上传器),但这些数据通常是罕见的。

    Validation at the trust boundaries is mostly present and in one place it is the wrong shape. The memory tool checks every scope write against an owner list, tools/manifest_gate.py rejects an ungrounded source with a non-zero exit rather than passing it through, the hooks bound their stdin reads, and the fetch path caps response size and redirect depth. The gap is that one tool accepts a caller-supplied URL and validates the scheme with a downstream denylist rather than an allowlist at the boundary, which this criterion asks for explicitly. The maintainer has the specifics; the fix is to allowlist the accepted schemes where the input enters and re-check after each redirect hop, with a regression test for the rejected cases.



    加固机制应该用于项目生产的软件,以便软件缺陷不太可能导致安全漏洞。 [hardening]
    加固机制可能包括HTTP头,如内容安全策略(CSP),用于减轻攻击的编译器标志(如-fstack-protector)或用以消除未定义行为的编译器标志。对于此条款的目的,最小权限不被认为是一种加固机制(最少权限是重要的,但是另有条款)。

    This criterion counts mechanisms such as HTTP response headers and compiler hardening flags, and explicitly excludes least privilege. The project produces no HTTP surface for a Content Security Policy to apply to and compiles nothing, so there is no -fstack-protector equivalent to enable. The protections it does use (bounded stdin reads, a response-size cap, a redirect cap, fail-closed gates and a credential-path denylist on agent reads) are real, but they are not what this criterion measures, so it is reported unmet rather than argued into a pass.



    该项目必须提供一个保证案例,证明其满足安全要求。保证案例必须包括:对威胁模型的描述,明确确定信任边界,确定设计原则得到适用,以及常见安全弱点已经被消减。 (需要网址) [assurance_case]
    一个保证案例是“一个文献记录的证据体系,提供了一个有说服力和有效的论据,指出一组关于系统属性的关键权利要求在给定环境中给定应用程序是充分合理的”(使用结构化保证案例模型的软件保证,Thomas Rhodes等人,NIST机构间报告7608)。信任边界是数据或执行改变其信任级别的边界,例如,典型Web应用程序中的服务器边界。常见做法是列出安全设计原则(例如Saltzer和Schroeer)和常见的实施安全漏洞(例如OWASP前10名或CWE / SANS前25名),并显示每个方案如何抵御。 BadgeApp保证案例可能是一个有用的例子。本条款与documentation_security,documentation_architecture和implement_secure_design等条款有关。

    https://github.com/cdeust/zetetic-team-subagents/blob/main/SECURITY.md covers two of the four required parts: it describes the threat (everything shipped executes in the user's session with no sandbox, and the payload does not even need to be compiled) and it argues that common implementation weaknesses are countered (shellcheck, CodeQL, Scorecard, the attested bundle and the install-path verifier). It does not identify trust boundaries as such and it makes no secure-design argument, so the assurance case is incomplete. Tracked with the four required parts as acceptance criteria in issue #70.


 分析 2/2

  • 静态代码分析


    如果至少有一个FLOSS工具能够以所选择的语言实现此条款,则该项目必须至少使用一个具有规则或方式的静态分析工具来查找分析语言或环境中的常见漏洞。 [static_analysis_common_vulnerabilities]
    专门设计用于寻找常见漏洞的静态分析工具更有可能找到它们。也就是说,使用任何静态工具通常会帮助找到一些问题,所以我们“通过”级别的徽章建议,但不要求这个条款。

    CodeQL runs the security-and-quality suite, which is the security query pack plus the quality queries, rather than a generic lint. It found real vulnerability-class defects and they were fixed: 33 alerts cleared at source in issue #60. https://github.com/cdeust/zetetic-team-subagents/blob/main/.github/workflows/codeql.yml


  • 动态代码分析


    如果由项目生成的软件包含使用内存不安全语言编写的软件(例如,C或C ++),项目必须至少使用一个动态工具(例如,fuzzer或web应用扫描程序)与一种检测缓冲区覆写等内存安全问题的机制组合例行使用。如果项目不生成以内存不安全语言编写的软件,请选择“不适用”(N/A)。 [dynamic_analysis_unsafe]
    检测内存安全问题的机制的示例包括AddressSanitizer(ASAN)(可在GCC和LLVM中使用),“Memory Sanitizer” valgrind 。其他可能使用的工具包括ThreadSanitizerUndefinedBehaviorSanitizer。广泛的断言也将起作用。

    The software is bash and Python. There is no C, C++ or other memory-unsafe code for a memory-safety dynamic analyser to target.



该数据可在社区数据许可协议 – 许可性,版本 2.0 (CDLA-Permissive-2.0)下获取。这意味着数据接收方可以共享数据,无论是否经过修改,只要数据接收方在共享数据时提供本协议文本。请注明Clement和OpenSSF最佳实践徽章贡献者。

项目徽章条目拥有者: Clement.
最后更新于 2026-07-27 22:32:34 UTC, 最后更新于 2026-07-27 22:58:32 UTC。 最后在 2026-07-27 22:58:12 UTC 获得通过徽章。