Description: This is a Skill review tool used to review the quality of other skills. The "incorrect examples" contained in the document (such as invalid YAML, incorrect naming, etc.) are for teaching demonstration only, showing how things should not be written. There is no malicious or obfuscated code.
Conducts a comprehensive review of skills according to Anthropic's official Skills development guide. Provides a structured review framework, scoring system, defect checklist, and improvement suggestions.
Core Responsibilities
- Structure Validation - Check folder structure and file naming conventions
Example:
ls skills/my-skill/CheckSKILL.mdwhether it exists - YAML Review - Validate frontmatter format and required fields
Example:
head -20 SKILL.mdCheck name/description - Description Evaluation - Check whether trigger conditions are clear
Example: description should include the "when the user..." trigger condition
- Organization Scoring - Evaluate whether the skill is organized by task and common operations are prioritized
Example: "## Encoding and Decoding" ✅ vs "## Theory" ❌
- Instruction Review - Evaluate the quality and completeness of the main body instructions
Example: are there workflows, examples, error handling
- Example Quality - Evaluate example density and executability
Example: 1 code block per 5-30 lines is the optimal density
- Tips Scoring - Evaluate the quality and value of the tips section
Example: 5-10 non-obvious practical tips
- Best Practices - Check compliance against the official guide
Example: check whether it is organized by task rather than abstract concepts
Workflow
Step 1: Receive review request
When the user requests a skill review:
- Confirm the skill folder path or read the contents of SKILL.md
- Determine the review strictness level:
Strict mode (must read the full official guide):
When the user expresses the following intentions, you must first read the references/anthropic-skills-development-guide.md full guide:
- Explicitly requests "strict check", "careful review", "comprehensive review"
- Mentions "high quality requirements", "production level", "pre-release check"
- Expresses "does not want any omissions", "according to the highest standards"
- Used for team/organization/company projects
- Preparing for public release or sharing with others
Regular mode:
- Prioritize reading
references/checklist.md(quick checklist) - Read the full official guide when encountering questions or edge cases
Step 2: Structure check
Folder structure validation:
[ ] 技能文件夹使用 kebab-case 命名(如 my-skill)
[ ] SKILL.md 存在且命名精确(区分大小写)
[ ] 无 README.md 在技能文件夹内
[ ] 可选文件夹(如存在)命名规范
Optional folder naming conventions (if present):
-
scripts/- All lowercase plural, no spaces/underscores -
references/- All lowercase plural, no spaces/underscores -
assets/- All lowercase plural, no spaces/underscores
❌ Incorrect example: Scripts、script、scripts_backup、References、refs、Assets、asset_files
Example: Check folder structure
# 查看技能文件夹结构
ls -la skills/china-holidays/
Expected output:
skills/china-holidays/
├── SKILL.md # ✅ 正确:精确命名
└── references/ # ✅ 正确:全小写复数
└── calendar-guide.md
Score: __/4
Step 3: YAML front matter check
Required field validation:
[ ] name 字段存在,kebab-case,无空格大写
[ ] description 字段存在,非空
[ ] YAML 分隔符完整(--- 开头和结尾)
[ ] 无 XML 标签(< >)
[ ] name 不以 claude/anthropic 开头
Example: Validate YAML front matter
# 读取前 20 行检查 YAML
head -20 skills/china-holidays/SKILL.md
Expected output (correct example):
---
name: china-holidays
description: 获取中国国家法定节假日安排。当用户询问"放假安排"、"节假日"时使用。
---
Incorrect example (should be rejected):
name: ChinaHolidays # ❌ 大写,应该 kebab-case
name: claude-scheduler # ❌ 以 claude 开头
description: "" # ❌ 空描述
<skill> # ❌ 包含 XML 标签
Description quality score (out of 8 points):
[2] 开头说明做什么(主动动词)
好:"分析 Figma 设计文件并生成开发者交接文档"
差:"这是关于 Figma 的技能"
[2] 包含触发条件("当用户...时")
好:"当用户上传 .fig 文件或询问设计规格时使用"
差:无触发条件
[2] 具体范围(提及具体工具、操作或场景)
好:"Figma 设计文件、.fig、组件文档、设计交接"
差:"帮助处理项目"
[2] 合理长度(50-200 字符,不超过 1024)
太短:无搜索价值
太长:被截断
Score: __/8
Step 4: Organization score
Organize by task/scenario (rather than by abstract concepts):
[2] 按任务/操作组织章节
好:"## 编码和解码" → "## 检查字符" → "## 转换格式"
差:"## 理论" → "## 类型" → "## 高级"
[2] 常见操作优先
好:基础用法 → 变体 → 高级 → 边界情况
差:配置说明 → 理论背景 → 最后才是基础用法
[1] 章节自包含(可独立使用)
[1] 深度一致(不混用 h2 与 h4 随机跳转)
Score: __/6
Step 5: Main body instruction check
Must include:
[ ] 清晰的工作流程或步骤说明
[ ] 至少一个使用示例
[ ] 错误处理或故障排查指南
Recommended to include (no point deduction):
[ ] 预期输出说明
[ ] 多个场景示例
[ ] 参考文档链接
[ ] "使用场景" 或 "When to Use" 部分
Score: __/6
Step 6: Example quality assessment
Example density calculation:
总行数:___
代码块数量:___
密度:每 ___ 行 1 个代码块
参考目标:每 5-30 行 1 个代码块
< 5 行/块:可能过于碎片化(短命令集或多命令速查除外)
> 40 行/块:需要更多示例
Example density score:
[3] 密度在 5-30 行/块范围内
[2] 密度略低(30-40 行/块)或略高(3-5 行/块)
[0] 密度严重不足(>40 行/块)或过高(<3 行/块)
Quality score for each example (0-3 points):
[ ] 语言标签正确(```bash, ```python 等)
[ ] 语法正确,命令可执行
[ ] 展示了预期输出或结果说明
[ ] 使用真实值(非 foo/bar/baz)
[ ] 无占位符(TODO, FIXME, xxx)
[ ] 自包含或有设置说明
0 分:broken 或有误导性
1 分:可用但简陋(无输出/上下文)
2 分:良好(正确,有输出或说明)
3 分:优秀(可直接复制,真实,覆盖边界)
Example quality score = average score of all examples × density score / 3 = __/9
Step 7: Executability assessment
Core question: Can the Agent produce correct results according to these instructions?
[3] 使用祈使句("运行 X"、"创建 Y")
非:"可以考虑..."或"建议..."或"You might..."
[3] 步骤顺序合理(前置条件在行动之前)
[2] 错误处理(说明失败时怎么做)
[2] 输出/结果描述(如何验证成功)
Score: __/10
Step 8: Progressive disclosure check
[ ] SKILL.md 控制在 500 行以内
500-600 行:给出提醒,不扣分
超过 600 行:扣分
[ ] 详细文档是否在 `references/` 中
[ ] 大文件(>300 行)有目录或结构说明
Score: __/3
Step 9: Tips score
Tips section quality assessment:
[2] 5-10 条技巧
少于 5 条:覆盖不足
多于 10 条:可能不够精炼
[2] 技巧非显而易见
好:"Makefile 头号陷阱:缩进必须用 Tab,不能用空格"
差:"确保测试你的代码"
[2] 技巧具体可执行
好:"使用 flock 防止 cron 任务重叠执行"
差:"小心并发执行"
[1] 技巧不矛盾主体内容
[1] 技巧覆盖特定主题的陷阱/踩坑点
Score: __/8
(Optional) Metadata reference
Check only when the skill contains a metadata field; for reference only, not scored
[ ] emoji 与技能主题相关
[ ] requires.anyBins 列出技能实际使用的工具(非 generic 如 bash)
[ ] os 数组准确(不包含不支持的平台)
[ ] JSON 格式有效
Score summary
SKILL 审核评分卡
═══════════════════════════════════════
技能名称:{name}
审核者:{agent/human}
日期:{date}
审核模式:{严格/常规}
类别 得分 满分
─────────────────────────────────────
结构检查 __ 4
Description 质量 __ 8
组织评分 __ 6
主体指令 __ 6
示例质量 __ 9
可执行性 __ 10
渐进式披露 __ 3
Tips 评分 __ 8
─────────────────────────────────────
总分 __ 54
转换为百分制:总分 × 1.85 = ___/100
评级:
85+ 优秀 — 可直接发布
70-84 良好 — 需要小改进
50-69 一般 — 需要明显改进
< 50 较差 — 需要重大修改
结论:{PUBLISH / REVISE / REWORK}
Common defects
Critical defects (block release)
缺陷:YAML 前置信息无效
检测:YAML 解析错误、缺少必填字段
修复:验证 YAML 格式,确保 name/description 都存在
缺陷:代码示例损坏
检测:语法错误、未定义变量、错误参数
修复:在干净环境中测试每个命令
缺陷:工具要求不匹配(如存在 metadata)
检测:requires 列出内容未在内容中使用
修复:grep 内容提取命令名,更新 requires 匹配
缺陷:误导性描述
检测:描述承诺的内容实际未覆盖
修复:使描述与实际内容一致,或补充缺失内容
Major defects (should be fixed before release)
缺陷:缺少"使用场景"部分
影响:Agent 不知道何时激活技能
修复:添加 4-8 条触发场景说明
缺陷:大段文字无示例
检测:任何超过 10 行无代码块的章节
修复:为每个概念添加具体示例
缺陷:示例缺少语言标签
检测:\`\`\` 后无语言标识符
修复:为每个代码块添加 bash/python/javascript/yaml 等
缺陷:缺少 Tips/技巧部分
影响:缺少使技能有价值的经验总结
修复:添加 5-10 条非显而易见的实用技巧
缺陷:按抽象概念组织
检测:章节名为"理论"、"概述"、"背景"、"介绍"
修复:按任务/操作重组:用户要做什么
Minor defects (recommended to fix)
缺陷:占位符值
检测:foo, bar, baz, example.com, TODO, FIXME
修复:替换为真实值
缺陷:格式不一致
检测:混合标题级别、代码块样式不一致
修复:统一标题层次和格式
缺陷:缺少交叉引用
检测:提到其他技能覆盖的工具/概念但未引用
修复:添加"参见 X 技能了解更多"注释
缺陷:过时的命令
检测:旧语法或已弃用工具
修复:更新为当前工具版本和语法
Quick review template
Quick review when a full score is not needed:
## 快速审核:{skill-name}
**结构**:[通过/问题:...]
**Description**:[强/弱:原因]
**示例**:[X 个代码块,共 Y 行 — 密度 正常/低/高]
**可执行性**:[Agent 可以/不可以 遵循这些指令,因为...]
**首要缺陷**:[最应该修复的单一问题]
**评分**:__/100
**结论**:[PUBLISH / REVISE / REWORK]
Review workflow
Self-check before release
# 1. 验证 YAML 前置信息
head -20 skills/my-skill/SKILL.md
# 目视确认 YAML 有效
# 2. 统计代码块数量
grep -c '```' skills/my-skill/SKILL.md
# 总行数除以这个数得密度
# 3. 检查占位符
grep -n -i 'todo\|fixme\|xxx\|foo\|bar\|baz' skills/my-skill/SKILL.md
# 4. 检查缺失语言标签
grep -n '^```$' skills/my-skill/SKILL.md
# 每个代码块都应该有语言标签
# 5. 验证工具要求匹配内容(如存在 metadata)
# 提取 requires,然后 grep 内容检查每个工具
# 6. 测试命令(抽样 3-5 个)
# 在干净 shell 中运行验证
# 7. 运行评分卡
# 目标:良好 35+,优秀 45+
Example: Complete review report output
## 审核报告:china-holidays
**结构**:✅ 通过 - kebab-case 命名,SKILL.md 存在
**Description**:✅ 强 - 包含触发条件和具体场景
**示例**:18 个代码块,524 行 — 密度 正常 (29 行/块)
**可执行性**:✅ 可以遵循 - 9 步清晰工作流程
**评分总结**:
─────────────────────────────────────
结构检查 4 4
Description 质量 8 8
组织评分 6 6
主体指令 6 6
示例质量 7 9
可执行性 10 10
渐进式披露 3 3
Tips 评分 8 8
─────────────────────────────────────
总分 52 54
百分制:96/100
**结论**:PUBLISH - 可直接发布
Review others' skills
# 安装技能(如适用)
npx molthub@latest install skill-name
# 阅读内容
cat skills/skill-name/SKILL.md
# 运行快速审核模板
# 如分数 < 25,考虑卸载并寻找替代
Reference files
Layered by usage scenario:
| File | Purpose | When to read |
|---|---|---|
references/checklist.md |
Quick checklist | Regular review process |
references/official-guide-summary.md |
Condensed guide | Quick reference for commonly used specifications |
references/anthropic-skills-development-guide.md |
Full official guide | Must be read in strict mode; consult when questions arise in regular mode |
Strict mode trigger conditions(triggered if any one is met):
- The user explicitly requests "strict check", "comprehensive review", "careful review"
- Mentions "high quality", "production level", "before release"
- Expresses "do not want any omissions", "according to the highest standards"
- Used for team/organization/company projects
- Preparing for public release or sharing
First determine the mode during review: In strict mode, the full guide must be read before reviewing; in regular mode, consult as needed.
Example
Example 1: Complete review
The user says: "Help me review this skill, the path is ./skills/china-holidays"
Action:
- Read
./skills/china-holidays/SKILL.md - Determine the review mode (regular by default)
- Check item by item according to the scorecard
- Generate review report
- Interpret results and provide improvement suggestions
Result: Output a complete review report, including a list of issues, revision suggestions, and a final score
Example 2: Strict mode review
User says: "This skill is about to be released for internal team use and needs strict checking, with no issues allowed"
Action:
- Trigger strict mode
- MustFirst read
anthropic-skills-development-guide.md - Strictly check item by item against the official guide
- Output a detailed report to ensure nothing is missed
Example 3: Quickly check the description
User says: "How is this skill's description written?" + attach the content
Action:
- Focus on analyzing the description field
- Use the 8-point scoring standard
- Provide improvement suggestions
Result: Point out issues and provide revision suggestions
Tips
-
Description is the most important — it accounts for more than 40% of actual impact. A perfect skill paired with a bad description will not be found.
-
Count code blocks first — a skill with fewer than 8 code blocks is almost always too abstract to be useful.
-
Test 3-5 commands in a clean environment — if more than 1 fails, the skill was not tested before release.
-
Organize by task vs organize by concept — this is the most critical structural quality difference. Good skills answer "how to do X"; bad skills explain "what X is."
-
A skill with good Tips but weak examples is more valuable than one with good examples but no Tips — Tips encode expertise that examples cannot convey.
-
Check whether requires matches actual usage — a common flaw is listing bash (which all have) instead of actual tools such as docker, curl, jq.
-
Skills that are too short (<150 lines) are usually not worth publishing — they provide less value than a quick web search. If a skill is too short, it may be a section of a larger skill.
-
Best standard: would you bookmark and use it yourself — if you would not use it yourself, do not publish it.