Token导航 LogoToken导航

Agent Skill · 开发

ears-authoring(耳朵创作)

用于辅助安全审计、权限检查、凭据风险、认证流程和常见漏洞排查。它适合让 Agent 梳理敏感配置、检查依赖风险、分析鉴权逻辑或生成安全复核清单。使用时不能把工具输出直接当最终结论,涉及密钥、令牌、用户数据或生产系统时,应先确认最小权限、脱敏方式和操作边界。

ears-authoring 是开发类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为需要联网;许可证信息还不明确,第三方扫描已通过,按安装量在开发类的 21,746 个 Skill 里排第 21,446。

总安装 188周安装 8GitHub Stars 61下载 66核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
unknown

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:ears-authoring(耳朵创作)
来源仓库:https://github.com/melodic-software/claude-code-plugins
仓库路径:skills/ears-authoring
安装命令:
npx skills add https://github.com/melodic-software/claude-code-plugins --skill ears-authoring
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/melodic-software/claude-code-plugins --skill ears-authoring

这个 Skill 是做什么的

开发类 Skill 面向写代码、改代码和排查问题,一般会读写仓库文件,部分还会执行构建或测试命令。

协助生成符合 EARS 语法规范的精确需求描述,提升需求可读性与执行一致性。适用于敏捷开发、产品迭代及跨团队沟通中对齐期望与减少歧义。输入原始需求后自动匹配最佳 EARS 模板并返回改写建议与适用理由说明。纯文本处理技能,无需外部依赖,但输出需经人工确认以避免误读业务逻辑。

ears-authoring 属于开发类 Skill,可作为该场景下的辅助能力补充。

来源仓库内路径:skills/ears-authoring

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 Codex、Claude、Cursor、Gemini CLI;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 npx skills add https://github.com/melodic-software/claude-code-plugins --skill e…;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它需要需要联网,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

ears-authoring 怎么安装?

在终端执行 npx skills add https://github.com/melodic-software/claude-code-plugins --skill ears-author…。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

ears-authoring 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

ears-authoring 需要什么权限?

权限级别是需要联网。该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

ears-authoring 是免费的吗?

公开资料没有标注明确许可证,收费和商用限制都无法确认。正式使用前请到来源仓库核对授权说明。

ears-authoring 用的人多吗?

公开数据是累计安装 188 次,近一周安装 8 次,下载 66 次,GitHub 61 星。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

ears-authoring 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。
  • 第三方扫描已通过,常见的高风险行为没有命中。

先想清楚的地方

  • 许可证没有明确标注,能不能商用、能不能再分发都不确定。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 Codex,想直接扩展它能力的人。
  • 需要处理开发类任务、又不想每次都手写提示词的人。

不太适合

  • 需要明确授权才能用到公司项目里的场景。
  • 需要厂商级技术支持和 SLA 的生产系统。

开发类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
skill-vetter

技能审查员 · OpenClaw

8,779,018986操作浏览器MIT-0通过
ontology

本体论 · OpenClaw

6,636,960576只读MIT-0通过
entra-app-registration

内部应用程序注册 · Codex、Claude

6,631,872764接触敏感数据未明确通过
entra-app-registration

内部应用程序注册 · Codex、Claude

3,757,680201接触敏感数据未明确通过
nano-banana-pro

纳米香蕉专业版 · OpenClaw

3,526,898364接触敏感数据MIT-0有提醒
ears-authoring

耳朵创作 · 本页

18861需要联网未明确通过

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

EARS Authoring

EARS (Easy Approach to Requirements Syntax) pattern authoring for precise, unambiguous requirements.

When to Use This Skill

Keywords: EARS patterns, ubiquitous requirements, state-driven requirements, event-driven requirements, unwanted behavior, optional features, complex requirements, SHALL keyword, requirement syntax

Use this skill when:

  • Writing new requirements using EARS syntax
  • Converting informal requirements to EARS format
  • Validating EARS pattern correctness
  • Selecting the right EARS pattern for a requirement
  • Understanding EARS anti-patterns

Quick Pattern Reference

PatternKeywordTemplate
Ubiquitous(none)The <entity> SHALL <action>
State-DrivenWHILEWHILE <condition>, the <entity> SHALL <action>
Event-DrivenWHENWHEN <trigger>, the <entity> SHALL <action>
UnwantedIF...THENIF <condition>, THEN the <entity> SHALL <action>
OptionalWHEREWHERE <feature>, the <entity> SHALL <action>
ComplexMultipleCombination of patterns

Pattern Selection Decision Tree

Start here: When does this requirement apply?

  1. Always applies (no conditions) → Use Ubiquitous: "The system SHALL..."
  2. While in a specific state → Use State-Driven: "WHILE in maintenance mode, the system SHALL..."
  3. When something happens (event/trigger) → Use Event-Driven: "WHEN user clicks submit, the system SHALL..."
  4. To handle unwanted behavior (error/exception) → Use Unwanted: "IF authentication fails, THEN the system SHALL..."
  5. Only when feature is enabled (optional/configurable) → Use Optional: "WHERE dark mode is enabled, the system SHALL..."
  6. Multiple conditions apply → Use Complex: "WHILE active, WHEN timeout occurs, the system SHALL..."

Pattern Details

Ubiquitous Pattern

Use when: Requirement applies unconditionally, always active.

Template:

The <entity> SHALL <action>

Keywords: None required (no WHILE, WHEN, IF, WHERE)

Examples:

  • "The system SHALL encrypt all data at rest"
  • "The API SHALL respond in JSON format"
  • "The application SHALL log all user actions"

Common Mistakes:

  • Adding unnecessary conditions when behavior is universal
  • Using "should" instead of "SHALL"

State-Driven Pattern

Use when: Behavior applies while system is in a particular state.

Template:

WHILE <condition>, the <entity> SHALL <action>

Keywords: WHILE (at start)

Examples:

  • "WHILE in maintenance mode, the system SHALL display a banner"
  • "WHILE the connection is active, the system SHALL send heartbeats"
  • "WHILE the user is authenticated, the system SHALL show the dashboard"

Common Mistakes:

  • Using WHEN instead of WHILE (WHEN = event, WHILE = state)
  • Describing events, not states

Event-Driven Pattern

Use when: Action triggered by a specific event or user action.

Template:

WHEN <trigger>, the <entity> SHALL <action>

Keywords: WHEN (at start)

Examples:

  • "WHEN a user submits the form, the system SHALL validate inputs"
  • "WHEN an error occurs, the system SHALL log the details"
  • "WHEN the session expires, the system SHALL redirect to login"

Common Mistakes:

  • Using WHILE instead of WHEN (WHILE = state, WHEN = event)
  • Describing states, not events

Unwanted Behavior Pattern

Use when: Handling exceptions, errors, or unwanted conditions.

Template:

IF <condition>, THEN the <entity> SHALL <action>

Keywords: IF...THEN (both required)

Examples:

  • "IF authentication fails, THEN the system SHALL lock the account"
  • "IF the database is unavailable, THEN the system SHALL queue requests"
  • "IF input validation fails, THEN the system SHALL display error messages"

Common Mistakes:

  • Using IF-THEN for normal behavior (use Event-Driven instead)
  • Missing THEN keyword
  • Using for positive conditions (reserve for negative/unwanted)

Optional Feature Pattern

Use when: Behavior depends on feature flag or configuration.

Template:

WHERE <feature/config>, the <entity> SHALL <action>

Keywords: WHERE (at start)

Examples:

  • "WHERE dark mode is enabled, the system SHALL use dark theme"
  • "WHERE audit logging is configured, the system SHALL log access"
  • "WHERE two-factor authentication is enabled, the system SHALL require OTP"

Common Mistakes:

  • Using IF instead of WHERE (IF = unwanted, WHERE = optional)
  • Describing mandatory features as optional

Complex Pattern

Use when: Requirement combines multiple conditions from different patterns.

Template:

<Pattern1>, <Pattern2>, the <entity> SHALL <action>

Examples:

  • "WHILE active, WHEN timeout occurs, the system SHALL reconnect"
  • "WHILE in production mode, IF error occurs, THEN the system SHALL notify ops"
  • "WHERE caching is enabled, WHEN data changes, the system SHALL invalidate cache"

Common Mistakes:

  • Using Complex when a simpler pattern suffices
  • Nesting conditions too deeply (max 2 recommended)

Writing Quality Requirements

The SHALL Keyword

Always use SHALL for requirements:

  • ✅ "The system SHALL validate input"
  • ❌ "The system should validate input"
  • ❌ "The system must validate input"
  • ❌ "The system will validate input"

Why: SHALL indicates mandatory behavior. Other words are ambiguous.

Active Voice

Always use active voice:

  • ✅ "The system SHALL encrypt data"
  • ❌ "Data shall be encrypted"
  • ❌ "Encryption shall be performed"

Single Requirement Per Statement

One action per requirement:

  • ✅ "The system SHALL validate input"
  • ✅ "The system SHALL log validation errors"
  • ❌ "The system SHALL validate input and log errors"

Testable Requirements

Requirements must be testable:

  • ✅ "The system SHALL respond within 200ms"
  • ❌ "The system SHALL respond quickly"
  • ✅ "The system SHALL support 1000 concurrent users"
  • ❌ "The system SHALL scale well"

Anti-Patterns to Avoid

Anti-PatternProblemFix
"should" instead of "SHALL"Ambiguous obligationUse "SHALL"
Passive voiceUnclear actorUse active voice
Multiple requirementsUntestable compoundSplit into separate
Implementation detailsSpecifies "how"Focus on "what"
Vague termsNot measurableAdd specific metrics
Wrong pattern keywordSemantic confusionMatch pattern to behavior

Validation Checklist

Before finalizing an EARS requirement:

  • Uses "SHALL" (not should/must/will)
  • Uses active voice (entity does action)
  • Single behavior per statement
  • Testable with measurable criteria
  • Pattern keyword matches behavior type
  • No implementation details (what, not how)

Integration with Canonical Spec

EARS requirements map to canonical specification:

requirements:
  - id: "REQ-001"
    text: "WHEN a user submits the form, the system SHALL validate inputs"
    priority: must
    ears_type: event-driven  # Matches pattern used
    acceptance_criteria:
      - id: "AC-001"
        given: "a user on the form page"
        when: "the user clicks submit"
        then: "the system validates all inputs"

References

Detailed Documentation:

Related Skills:

  • canonical-spec-format - Canonical specification structure
  • spec-management - Specification workflow navigation
  • requirements-quality - INVEST criteria and quality assessment

Last Updated: 2025-12-24

Version History

  • v1.0.0 (2025-12-26): Initial release

平台分布

Codex

34.86%
按下载量换算23

Claude

30.63%
按下载量换算20

Cursor

18.65%
按下载量换算12

Gemini CLI

10.49%
按下载量换算7

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

需要联网

该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills