Token导航 LogoToken导航TokenDH.com
待分类需要联网github未标认证来源可访问clear审计通过

aeo-spec-validatoraeo 规范验证器

Agent Skill

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。

总安装

321

周安装

13

GitHub Stars

公开资料未说明

下载量

101
CodexClaudeCursorGemini CLI

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

3

许可证

MIT

最后核验

2026-05-01

来源状态

来源可访问

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:aeo-spec-validator(aeo 规范验证器)
来源仓库:https://github.com/ivzc07/aeo-skills
仓库路径:skills/aeo-spec-validator
安装命令:
npx skills add https://github.com/ivzc07/aeo-skills --skill aeo-spec-validator
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。不同来源提供的安装方式可能略有差异;本站展示可直接复制的安装命令,安装前请核对来源页面。

skills.shnpx skills
npx skills add https://github.com/ivzc07/aeo-skills --skill aeo-spec-validator

简介

AEO 规范验证器用于评估任务描述的清晰度与完整性,评分 0-100 分。

  • 适用于启动新任务前判断需求是否明确、验收标准是否可测。
  • 提供客观性、接受条件、依赖项等维度的打分指引与改进建议。
  • 不能替代人工判断,需结合实际业务背景复核其输出结论。
  • aeo-spec-validator 属于待分类类 Skill,可作为该场景下的辅助能力补充。

SKILL.md

AEO Spec Validator

Purpose: Validate task specifications and score them 0-100 to ensure they're sufficiently defined before execution.

When to Use

Invoke this skill when:

  • User provides a task description
  • Before starting any implementation work
  • When task is ambiguous or unclear

Scoring System (0-100)

Clarity Indicators (0-50 points)

Objective Clarity (15 points):

  • 15 pts: Explicit, unambiguous objective

- Example: "Add email validation to signup form with regex check"

  • 10 pts: Clear but minor ambiguities

- Example: "Add email validation to signup"

  • 5 pts: Vague objective

- Example: "Improve signup process"

  • 0 pts: No clear objective

Acceptance Criteria (15 points):

  • 15 pts: Specific, testable criteria defined

- Example: "Must validate RFC 5322 format, reject + aliases, show inline errors"

  • 10 pts: General criteria mentioned

- Example: "Must validate email format and show errors"

  • 5 pts: Implied criteria

- Example: "Should work for valid emails"

  • 0 pts: No criteria mentioned

Context Provided (10 points):

  • 10 pts: Full context (where, why, constraints)

- Example: "For the signup form in /pages/auth/signup.tsx, using existing validator utility"

  • 5 pts: Partial context

- Example: "For the signup form"

  • 0 pts: No context

Dependencies Identified (10 points):

  • 10 pts: All dependencies listed (libraries, services, APIs)

- Example: "Uses validator.js library, calls POST /api/validate"

  • 5 pts: Some dependencies mentioned

- Example: "Uses validator library"

  • 0 pts: No dependencies mentioned

Quality Indicators (0-30 points)

Tech Stack Specified (10 points):

  • 10 pts: Specific technologies/libraries named
  • 5 pts: General tech mentioned (e.g., "use a validation library")
  • 0 pts: No tech mentioned

Test Requirements (10 points):

  • 10 pts: Test cases specified

- Example: "Test valid emails, invalid formats, edge cases (+ alias)"

  • 5 pts: General testing mentioned

- Example: "Should have tests"

  • 0 pts: No testing mentioned

Performance/Security (10 points):

  • 10 pts: Explicit requirements

- Example: "Must reject in <100ms, prevent email injection"

  • 5 pts: General concerns mentioned

- Example: "Should be fast and secure"

  • 0 pts: No mention

Risk Assessment (0-20 points)

Scope (10 points):

  • 10 pts: Small, well-defined scope (1-2 files, <100 LOC)
  • 5 pts: Medium scope (3-5 files, 100-500 LOC)
  • 0 pts: Large scope (5+ files, 500+ LOC)

Complexity (10 points):

  • 10 pts: Simple CRUD or clear logic
  • 5 pts: Moderate complexity (multiple systems, integration)
  • 0 pts: High complexity (architectural changes, unknowns)

Ambiguity Detection

Automatically detect and penalize these red flags:

Subjective Terms (-5 each):

  • "fast", "quick", "performant"
  • "good", "better", "optimal"
  • "simple", "clean", "elegant"
  • "user-friendly", "intuitive"

Vague Verbs (-10 each):

  • "improve" (without specifics)
  • "optimize" (without metrics)
  • "enhance" (without details)
  • "fix" (without describing what's broken)

Missing Constraints (-5 each):

  • No performance requirements
  • No error handling specified
  • No edge cases mentioned
  • No security considerations (for sensitive areas)

Scoring Examples

Example 1: Well-Defined Spec (Score: 92/100)

Task: Add email validation to the signup form in /pages/auth/signup.tsx

Requirements:
- Validate using RFC 5322 format via validator.js library
- Reject email addresses with + aliases
- Show inline error message "Invalid email format" on blur
- Call POST /api/validate-email to check if already registered
- Tests: valid emails, invalid formats, + alias rejection, duplicates

Score Breakdown:
- Objective clarity: 15/15 (explicit)
- Acceptance criteria: 15/15 (specific, testable)
- Context: 10/10 (file location, existing utility)
- Dependencies: 10/10 (validator.js, API endpoint)
- Tech stack: 10/10 (validator.js named)
- Test requirements: 10/10 (specific test cases)
- Performance/security: 7/10 (missing perf req)
- Scope: 10/10 (single file)
- Complexity: 5/10 (integration but clear)

Total: 92/100 → PROCEED

Example 2: Poor Spec (Score: 28/100)

Task: Improve the signup

Score Breakdown:
- Objective clarity: 5/15 (vague)
- Acceptance criteria: 0/15 (none)
- Context: 0/10 (none)
- Dependencies: 0/10 (none)
- Tech stack: 0/10 (none)
- Test requirements: 0/10 (none)
- Performance/security: 0/10 (none)
- Scope: 8/10 (assume small)
- Complexity: 5/10 (assume simple)

Ambiguity Penalties:
- "Improve" (vague verb): -10
- "signup" (subjective good?): 0

Total: 28/100 → REFUSE

Feedback:
❌ SPEC INSUFFICIENT (28/100)

Missing:
• Specific acceptance criteria
• What to improve about signup?
• Context (which signup flow?)
• Dependencies and tech stack
• Test requirements

Please provide:
1. What specific improvement is needed?
2. Acceptance criteria for "done"
3. Which signup form/page?
4. Any constraints or requirements

Actions by Score Range

80-100: PROCEED

  • Well-defined spec
  • Proceed with confidence calculation
  • Note: "Spec score: XX/100 - well-defined"

60-79: MINOR GAPS

  • Generally clear, some details missing
  • Proceed but note assumptions
  • Format: ⚠️ SPEC HAS MINOR GAPS (68/100) Assumptions: • Using existing test framework • Standard error handling • No special performance requirements Proceeding with these assumptions. Correct if wrong.

40-59: MAJOR GAPS

  • Significant ambiguities
  • Ask clarifying questions before proceeding
  • Format: ❌ SPEC NEEDS CLARIFICATION (45/100) Missing Details: • Which file(s) should be modified? • What validation library to use? • Acceptance criteria not specified • No test requirements Please clarify: 1. Where should this be implemented? 2. What tech stack/libraries? 3. What defines "done"?

< 40: UNACCEPTABLE

  • Too vague to execute
  • Refuse task
  • Request complete spec
  • Format: ❌ CANNOT PROCEED - SPEC TOO UNCLEAR (28/100) This task is too vague. Please provide: 1. **Objective:** What exactly needs to be done? 2. **Acceptance Criteria:** How do we know it's done? 3. **Context:** Where/why is this needed? 4. **Dependencies:** What libraries/services? Example of a good spec: "Add email validation to /pages/auth/signup.tsx using validator.js. Must validate RFC 5322 format, reject + aliases, show inline errors. Tests for valid, invalid, and duplicate emails."

Integration Flow

  1. Invoke: Called by aeo-core during Phase 0
  2. Score: Analyze task and calculate score
  3. Return: Score + feedback if needed
  4. Decision: aeo-core uses score for confidence calculation

Examples for Reference

Show these examples if user asks for clarification:

Good Spec Template

[Task Name]

**Objective:** [Specific action to take]

**Location:** [File paths, components, modules]

**Requirements:**
- [Requirement 1]
- [Requirement 2]

**Dependencies:**
- Libraries: [list]
- Services: [list]
- APIs: [list]

**Acceptance Criteria:**
- [Criteria 1 - testable]
- [Criteria 2 - testable]

**Tests:**
- [Test case 1]
- [Test case 2]

**Constraints:**
- Performance: [requirements]
- Security: [requirements]

Common Mistakes

❌ "Make it faster" ✅ "Reduce API response time from 2s to <500ms by adding caching"

❌ "Fix the bug" ✅ "Fix null reference error in UserService.getUser() when user ID not found"

❌ "Add authentication" ✅ "Add JWT authentication to /api/* routes using bcrypt for password hashing"

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

需要对比不同来源的安装命令和来源信息时

04

需要参考平台分布和安装热度时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

补充不同宿主或平台的使用分布数据

能力 5

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

Claude Code

26.33%
按下载量换算27

windsurf

20.99%
按下载量换算21

Codex

16.11%
按下载量换算16

OpenCode

14.02%
按下载量换算14

trae

8.44%
按下载量换算9

Antigravity

3.74%
按下载量换算4

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。

来源信息

继续浏览同类 Skills