Token导航 LogoToken导航

Agent Skill · 研究检索

content-editing(内容编辑)

用于辅助文档、README、Markdown、说明文和内容稿件的整理与改写。它适合让 Agent 提炼结构、补齐章节、统一术语、检查链接或把零散材料整理成可读文档。使用时应保留项目已有事实、命令和路径,不要把未确认的信息写成确定结论;涉及对外文案时,还需要控制语气,避免过度营销或夸大能力。

content-editing 是研究检索类 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,按安装量在研究检索类的 73,675 个 Skill 里排第 71,089。

总安装 196周安装 8GitHub Stars 1下载 63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:content-editing(内容编辑)
来源仓库:https://github.com/azeem-2/final-docusarus
仓库路径:skills/content-editing
安装命令:
npx skills add https://github.com/azeem-2/final-docusarus --skill content-editing
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/azeem-2/final-docusarus --skill content-editing

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

用于系统性审查和改进书籍稿件,使其达到专业出版标准。适用于完成稿件的语法、风格、机械错误和事实准确性检查。提供六阶段编辑流程,包括结构审查、术语一致性和多轮审校建议。

安装方式:通过 GitHub 仓库安装,命令为 npx skills add https://github.com/azeem-2/final-docusarus --skill content-editing。注意:需用户提供完整草稿,重点在于质量提升而非创意生成。

来源仓库内路径:skills/content-editing

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/azeem-2/final-docusarus --skill content-editin…;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

content-editing 怎么安装?

在终端执行 npx skills add https://github.com/azeem-2/final-docusarus --skill content-editing。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

content-editing 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。

content-editing 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。

content-editing 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。

content-editing 用的人多吗?

公开数据是累计安装 196 次,近一周安装 8 次,下载 63 次,GitHub 1 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。

content-editing 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。
  • 需要处理研究检索类任务、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。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
content-editing

内容编辑 · 本页

1961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

Content Editing Skill

Systematic procedures for reviewing and improving book content to meet professional publication standards.

When to Use This Skill

  • Reviewing completed drafts for quality
  • Checking grammar, style, and mechanics
  • Verifying factual accuracy and citations
  • Ensuring terminology and voice consistency
  • Preparing content for final formatting
  • Conducting multi-pass editorial reviews

Editorial Workflow (Six-Pass System)

Pass 1: Structural Review (Macro Level)

Focus: Document organization and architecture

Checklist:

  • Document follows outline structure
  • All sections have clear, distinct purposes
  • Logical flow between sections
  • Appropriate section lengths (no extreme outliers)
  • Heading hierarchy consistent (H2 → H3 → H4, no skips)
  • Chapter/section breakdown matches plan
  • TOC-worthy headings properly formatted

Questions to Ask:

  • Does each section serve a clear purpose?
  • Are topics in the most logical order?
  • Are any sections too long and need splitting?
  • Are any sections too short and should be combined?

Pass 2: Content Quality (Micro Level)

Focus: Sentence and paragraph quality

Grammar & Mechanics:

  • Subject-verb agreement correct
  • Verb tense consistent within sections
  • Pronoun antecedents clear
  • No sentence fragments (unless intentional)
  • No run-on sentences
  • No comma splices
  • Proper punctuation throughout

Clarity & Concision:

  • Every sentence has clear meaning
  • No unnecessary words or phrases
  • Complex ideas broken into manageable chunks
  • Technical jargon defined or eliminated
  • Ambiguous pronouns resolved

Style & Voice:

  • Active voice predominates (target: >80%)
  • Consistent perspective (you/we/one)
  • Consistent tone (formal/balanced/conversational)
  • Parallel structure in lists
  • Varied sentence structure for rhythm

Readability Assessment (dynamic based on content type):

  • Flesch Reading Ease appropriate for audience (see table below)
  • Average sentence length matches complexity level
  • Paragraph lengths appropriate (4-6 sentences typical)
Content TypeTarget FleschSentence Length
General audience65-8012-18 words
Business/professional60-7015-20 words
Introductory technical55-6515-22 words
Advanced technical35-5518-28 words
Specialized (ML, compilers)30-5020-30 words
📝 Note: Lower Flesch scores are acceptable for advanced topics. The goal is clarity for the intended audience, not a universal threshold.

Pass 3: Citation Verification (Accuracy Check)

Focus: Citation completeness and accuracy

Citation Coverage:

  • Every factual claim has citation
  • Every statistic cited
  • Every direct quote cited
  • Every date/historical fact cited
  • Every technical specification cited

Citation Format:

  • Consistent format throughout (APA/MLA/Chicago)
  • In-text citations match reference list
  • All citation elements present (author, year, title, source, URL, access date)
  • Page numbers for quotes and specific claims

Citation Quality:

  • Sources are authoritative (academic, reputable news, official docs)
  • Sources are current (< 5 years unless historical)
  • Multiple sources for major claims
  • Primary sources used where possible

Cross-Reference with Research:

  • Every cited source exists in research notes
  • Source details match research documentation
  • No "dead" citations (source cannot be located)

Pass 4: Consistency Audit (Uniformity Check)

Focus: Terminology, style, and format uniformity

Terminology Consistency (Use Grep to identify variations):

# Check AI terminology consistency
grep -E "(AI|artificial intelligence|Artificial Intelligence|A.I.)" content/

# Check website spelling
grep -E "(web site|website|web-site)" content/

# Check hyphenation patterns
grep -E "(e-mail|email|Email)" content/

Create Terminology Glossary:

TermApproved FormAvoid
Artificial IntelligenceAI (after first mention)A.I., artificial intelligence (subsequent)
Websitewebsiteweb site, web-site
Emailemaile-mail, E-mail

Style Guide Compliance:

  • Number style consistent (spell out 1-10 or always numerals)
  • Date format consistent (Month Day, Year)
  • Time format consistent (12-hour with am/pm or 24-hour)
  • Oxford comma usage consistent
  • Capitalization consistent (job titles, headings)
  • Hyphenation consistent (compound modifiers, prefixes)

Voice & Perspective:

  • Same perspective throughout (you/we/one)
  • Same tone level maintained
  • Same formality level

Pass 5: Visual Formatting Audit

Focus: Ensuring content is visually engaging and properly formatted

📝 Note: Visual formatting rules are contextual, not rigid. Technical content benefits from more frequent visual breaks; conceptual/philosophical content may flow better with fewer interruptions. The goal is cognitive clarity, not arbitrary quotas.

Text Wall Detection (contextual):

  • Visual breaks inserted where they aid comprehension
  • Technical content: more frequent breaks (code, tables, lists)
  • Conceptual content: longer prose passages acceptable if flow is maintained
  • Long explanations broken up with lists, callouts, or code blocks where helpful
  • White space used effectively between logical groups

Callout Box Verification:

  • Appropriate callouts per section (2-4 for technical, fewer for conceptual)
  • Callout types match content (💡 for insights, ⚠️ for warnings, etc.)
  • Callouts not overused (losing impact)
  • Callout formatting consistent throughout

Code Block Quality (for technical content):

  • All code blocks have language specification (python,javascript, etc.)
  • Code includes explanatory comments
  • Code is properly indented and formatted
  • No orphaned code without surrounding explanation

Table Usage:

  • Tables used for comparisons and structured data
  • Table headers are clear and descriptive
  • Table formatting consistent (alignment, borders)
  • No tables with only 1-2 rows (use list instead)

List Formatting:

  • Bullet points for unordered items
  • Numbered lists for sequential steps
  • Parallel grammatical structure in list items
  • List items have consistent punctuation

Emphasis Consistency:

  • Bold used for key terms and emphasis
  • *Italics* used for foreign terms and subtle emphasis
  • inline code used for technical terms and commands
  • No mixing of emphasis styles for same purpose

Visual Diagram Review (where applicable):

  • ASCII diagrams properly aligned
  • Diagrams have clear labels
  • Complex relationships visualized, not just described

Section Separators:

  • Horizontal rules (---) used between major sections
  • Visual breathing room between topics
  • Consistent separator usage throughout

Common Visual Issues to Flag:

⚠️ TEXT WALL: [location] - Dense passage may benefit from visual break (consider context)
⚠️ MISSING CALLOUT: [section] - Key insight not highlighted (if appropriate for content type)
⚠️ UNSPECIFIED CODE: [line] - Code block missing language
⚠️ POOR TABLE: [location] - Consider converting to list
⚠️ INCONSISTENT EMPHASIS: [term] - Bold in some places, not others
⚠️ OVER-FORMATTED: [section] - Too many visual breaks disrupting narrative flow

Pass 6: Factual Accuracy (Truth Verification)

Focus: Fact-checking against research sources

Verification Requirements:

  • All statistics match source data exactly
  • All quotes are verbatim (no paraphrasing in quotes)
  • All dates and names spelled correctly
  • All technical specifications accurate
  • No unsupported generalizations

Cross-Reference Method:

  1. Identify claim in draft
  2. Locate cited source in research notes
  3. Verify claim matches source exactly
  4. Check for context (is claim misrepresented?)
  5. Flag discrepancies for writer review

Confidence Levels:

  • High: 3+ authoritative sources agree
  • Medium: 2 sources agree, or 1 highly authoritative source
  • Low: Single source of moderate authority
  • Unverified: No source found or sources conflict

Flag Low/Unverified claims for additional research.

Grammar and Style Rules Reference

Common Grammar Errors

Subject-Verb Agreement:

  • Wrong: "The team of editors review drafts."
  • Correct: "The team of editors reviews drafts."

Comma Splices:

  • Wrong: "The edit is complete, the draft is ready."
  • Correct: "The edit is complete; the draft is ready."

Misplaced Modifiers:

  • Wrong: "She only edited three chapters."
  • Correct: "She edited only three chapters."

Style Preferences (Configurable per Project)

Numbers: Spell out 1-10, numerals for 11+ (or choose consistent alternative) Dates: Month Day, Year (January 15, 2025) Oxford Comma: Choose one style and apply consistently Hyphenation: Compound modifiers before noun hyphenate, after noun no hyphen Capitalization: Job titles capitalize before name, lowercase after

Readability Metrics

Flesch Reading Ease Score

Score Interpretation:

  • 90-100: Very Easy (5th grade) - Children's books
  • 80-89: Easy (6th grade) - Conversational writing
  • 70-79: Fairly Easy (7th grade) - General audience
  • 60-69: Standard (8th-9th grade) - Business writing
  • 50-59: Fairly Difficult (10th-12th grade) - Academic
  • 40-49: Difficult - Advanced technical
  • 30-39: Very Difficult - Highly specialized (compilers, physics, ML theory)
  • Below 30: Expert only - Mathematical proofs, research papers

Target for Book Generation (dynamic by content type):

Content TypeTarget FleschAcceptable Range
General audience65-8060-85
Business/professional60-7055-75
Introductory technical55-6550-70
Intermediate technical45-6040-65
Advanced technical35-5530-60
Specialized/theoretical30-5025-55
⚠️ Warning: Do NOT force higher Flesch scores on advanced technical content. Simplifying specialized terminology can reduce precision and accuracy. The goal is appropriate clarity for the intended audience.

When to Improve Score (general audience only):

  • Shorten sentences (< 20 words average)
  • Use simpler words (fewer syllables)
  • Break complex sentences into multiple sentences
  • Replace jargon with plain language

When NOT to Simplify (technical content):

  • Technical terminology is necessary for precision
  • Audience expects domain vocabulary
  • Simplification would lose meaning or accuracy

Passive Voice Percentage

Target: < 20%

Detection Pattern: [form of "to be"] + [past participle]

How to Fix:

  1. Identify actor: WHO performs the action?
  2. Rewrite with actor as subject

- Passive: "The draft was reviewed by the editor." - Active: "The editor reviewed the draft."

Edit Summary Report Template

# Edit Summary: [Document Title]

**Edited**: [YYYY-MM-DD]
**Word Count**: [original] → [revised] ([+/- change])
**Total Changes**: [number]

## Changes by Category

### Grammar & Mechanics: [count]
### Citations: [count]
### Clarity & Flow: [count]
### Consistency: [count]
### Visual Formatting: [count]
### Factual Corrections: [count]

## Quality Metrics

| Metric | Before | After | Target | Status |
|--------|--------|-------|--------|--------|
| Flesch Reading Ease | [score] | [score] | [per content type] | ✅/❌ |
| Citation Coverage | [%] | [%] | 100% | ✅/❌ |
| Passive Voice % | [%] | [%] | <20% | ✅/❌ |
| Avg Sentence Length | [words] | [words] | [per content type] | ✅/❌ |
| Text Walls Fixed | [count] | 0 | 0 | ✅/❌ |
| Callouts Per Section | [avg] | [avg] | 2-4 | ✅/❌ |
| Code Blocks w/ Language | [%] | [%] | 100% | ✅/❌ |

**Content Type**: [general/business/intro-tech/advanced-tech/specialized]

## Visual Formatting Summary
- **Callouts added**: [count] (💡: [n], ⚠️: [n], 🎯: [n], etc.)
- **Tables created**: [count]
- **Lists converted**: [count] (from run-on sentences)
- **ASCII diagrams added**: [count]
- **Text walls broken up**: [count]

## Issues Requiring Author Review
- [Unverifiable claims, technical accuracy questions]

## Recommendations
- [Patterns to watch in future writing]

Common Editing Pitfalls

  1. Over-Editing: Edit for correctness and clarity, not personal preference
  2. Missing Context: Read surrounding paragraphs before making changes
  3. Inconsistent Application: Use Grep to find ALL instances, apply rule uniformly
  4. Citation Overload: Distinguish common knowledge from factual claims
  5. Ignoring Readability: Balance formal correctness with reader comprehension

Quality Assurance

Before marking edit complete:

  • All six editorial passes completed
  • 100% citation coverage verified
  • Readability target achieved (Flesch > 60)
  • Consistency issues resolved
  • Visual formatting standards met
  • Edit summary report generated
  • Flagged issues documented for author

Skill Version: 1.3.0 Last Updated: 2025-11-27 Maintained By: Universal Pedagogical Engine Team

平台分布

Codex

38.68%
按下载量换算24

Claude

26.7%
按下载量换算17

Cursor

18.42%
按下载量换算12

Gemini CLI

9.01%
按下载量换算6

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