Token导航 LogoToken导航

Agent Skill · 研究检索

seo-content(SEO 内容)

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。

seo-content 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,第三方扫描有提醒项,按安装量在研究检索类的 73,675 个 Skill 里排第 70,604。

总安装 210周安装 9GitHub Stars 28下载 73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:seo-content(SEO 内容)
来源仓库:https://github.com/mangollc/claude-seo-skill
仓库路径:skills/seo-content
安装命令:
npx skills add https://github.com/mangollc/claude-seo-skill --skill seo-content
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/mangollc/claude-seo-skill --skill seo-content

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。

用于辅助文档、README、Markdown 和内容稿件的整理与改写。适合让 Agent 提炼结构、补齐章节、统一术语或检查链接有效性。通过 npx skills add 命令从指定 GitHub 仓库安装,需保留项目已有事实和路径。不要把未确认的信息写成确定结论,对外文案应控制语气避免过度营销。

使用前应评估是否会触发文件修改,确保内容准确性和一致性。

来源仓库内路径:skills/seo-content

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/mangollc/claude-seo-skill --skill seo-content;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

seo-content 怎么安装?

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

seo-content 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。

seo-content 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。

seo-content 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。

seo-content 用的人多吗?

公开数据是累计安装 210 次,近一周安装 9 次,下载 73 次,GitHub 28 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。

seo-content 安全吗?

第三方扫描有提醒项,说明存在需要人工确认的行为。安装前请打开源码看清它读了什么、联网到哪里。

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。

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。
  • 第三方扫描存在提醒项或未通过项,需要人工逐条确认。
  • 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
seo-content

SEO 内容 · 本页

21028需要联网未明确有提醒

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 Optimization

Load client context first. See seo-keywords rules/client-setup.md.

File Organization

content/
├── page-scores/
│   └── {url-slug}-score.json      # Individual page scores
├── content-gaps.json              # Gap analysis
├── content-decay.json             # Decay detection results
├── cannibalization.json           # Pages competing for same keywords
├── pre-publish-checklist.md       # Reusable checklist template
└── serp-previews/
    └── {url-slug}-preview.md      # SERP preview renders

On-Page SEO Scoring (0-100)

Scrape the target page:

firecrawl scrape {url} --format markdown,links -o .seo/clients/{slug}/content/.firecrawl/page-{slug}.json
firecrawl scrape {url} --html -o .seo/clients/{slug}/content/.firecrawl/page-{slug}.html

Scoring Rubric

CategoryPointsChecks
Title Tag15Length 50-60 chars (deduct 5 if outside range), primary keyword in first half (5 pts), unique across site (5 pts), SERP truncation safe (5 pts)
Meta Description10Length 120-155 chars (deduct 5 if outside), contains keyword (3 pts), has CTA/action word (4 pts), unique (3 pts)
H1 Tag10Exactly one H1 (5 pts), contains primary keyword (5 pts)
Heading Hierarchy10Proper H1>H2>H3 nesting with no skipped levels (5 pts), keywords in at least 2 subheadings (3 pts), sufficient subheadings for content length (2 pts)
Content Depth15Word count >= avg of top 5 competitors (8 pts), covers all major subtopics (7 pts)
Keyword Usage10Density 1-2% (4 pts), keyword in first 100 words (3 pts), LSI/semantic variations used (3 pts)
Internal Links10Min 3 internal links (4 pts), relevant anchor text (3 pts), links to pillar/related content (3 pts)
Image Optimization10All images have alt text (4 pts), descriptive filenames (2 pts), modern format WebP/AVIF (2 pts), lazy loading on below-fold (2 pts)
URL Structure5Short slug < 60 chars (2 pts), contains keyword (2 pts), no parameters/IDs (1 pt)
Social Meta5og:title + og:description + og:image present (3 pts), twitter:card present (2 pts)

How to Check Each

Title tag: Extract <title> from HTML. Count characters. Check keyword position.

Meta description: Extract <meta name="description">. Count characters. Look for action words (discover, learn, get, find, try, start, compare).

H1 tag: Count <h1> occurrences. Should be exactly 1. Check keyword presence.

Heading hierarchy: List all headings in order. Verify H2 follows H1, H3 follows H2. Flag skipped levels (H1 → H3).

Content depth: Count words in main content (exclude nav, footer, sidebar). Compare against top 5 competitors for the same keyword:

firecrawl search "{target-keyword}" --scrape --limit 5 -o .seo/clients/{slug}/content/.firecrawl/competitors-{slug}.json --json

Keyword usage: Count occurrences of primary keyword. Calculate density = (count / total words) * 100. Check if keyword appears in first 100 words. Look for semantic variations and LSI terms.

Internal links: From the links data, count links to same domain. Check anchor text diversity.

Images: Parse all <img> tags. Check alt attribute, src filename, loading attribute, image format.

URL: Analyze URL path. Check length, keyword presence, parameter presence.

Social meta: Check for og: and twitter: meta tags in HTML head.

SERP Preview Generator

Show how the page will appear in Google search results:

┌─────────────────────────────────────────────────┐
│ {title_tag_truncated_at_60_chars}...            │
│ {display_url}                                    │
│ {meta_description_truncated_at_155_chars}...     │
│                                                  │
│ Rich Results: [FAQ ✓] [Review Stars ✗] [Video ✗]│
└─────────────────────────────────────────────────┘

Check:

  • Title truncation: will Google cut it off? (> 60 chars or > 580px pixel width)
  • Description truncation: > 155 chars gets cut
  • URL display: clean and readable?
  • Rich result eligibility: based on schema markup present

Content Readability Analysis

Calculate readability metrics from the page content:

  • Flesch-Kincaid Grade Level: target 7-9 for most web content

- Formula: 0.39 * (words/sentences) + 11.8 * (syllables/words) - 15.59

  • Average sentence length: target 15-20 words
  • Paragraph length: target 2-4 sentences per paragraph
  • Passive voice usage: flag if > 10% of sentences
  • Transition word usage: should appear in 30%+ of sentences

Provide grade: A (grade 6-8), B (grade 9-10), C (grade 11-12), D (grade 13+).

E-E-A-T Analysis

Based on Google's Quality Rater Guidelines, evaluate:

Experience (E1)

  • Does the content show first-hand experience?
  • Look for: personal anecdotes, original photos, specific details only someone with experience would know
  • Score: Strong / Moderate / Weak / Missing

Expertise (E2)

  • Does the author demonstrate subject expertise?
  • Check for: author byline, credentials, author page, depth of knowledge in content
  • Score: Strong / Moderate / Weak / Missing

Authoritativeness (A)

  • Is this source/author recognized as an authority?
  • Check for: brand mentions in other sources, citations, industry recognition, domain age
  • Score: Strong / Moderate / Weak / Missing

Trustworthiness (T)

  • Can users trust this content and site?
  • Check for: HTTPS, contact information, privacy policy, terms of service, sources cited in content, factual accuracy indicators
  • Score: Strong / Moderate / Weak / Missing

E-E-A-T recommendations should be specific: "Add author bio with credentials" not "improve E-E-A-T."

Featured Snippet Optimization

Analyze target keyword for featured snippet opportunity:

firecrawl search "{keyword}" --scrape --limit 3 -o .seo/clients/{slug}/content/.firecrawl/snippet-{slug}.json --json

Snippet types and how to win them:

TypeFormat RequiredExample
ParagraphDirect answer in 40-60 words, immediately after an H2 matching the question"What is [topic]?" → 2-3 sentence definition
ListOrdered/unordered list with 5-8 items, preceded by H2 with question"How to [do thing]" → numbered steps
TableHTML <table> with clear headers, structured comparison data"Comparison of [X] vs [Y]" → feature comparison

Optimization steps:

  1. Check if a featured snippet currently exists for the target keyword
  2. Identify the snippet type (paragraph, list, table)
  3. Recommend content format changes to match
  4. Provide a draft snippet-optimized content block

Content Gap Analysis

Compare client content against top competitors for a topic cluster:

# Scrape competitor content for the same topic
firecrawl search "{topic} site:{competitor-domain}" --scrape --limit 10 -o .seo/clients/{slug}/content/.firecrawl/comp-gap-{slug}.json --json

Identify:

  • Subtopics competitors cover that client does not
  • Content formats competitors use (video, infographic, interactive, calculator)
  • Questions competitors answer that client doesn't address
  • Depth differences (word count, number of examples, data points)

Output as prioritized content brief suggestions.

Content Decay Detection

Identify content that needs refreshing:

Decay signals (check for each published page):

  1. Outdated statistics: years mentioned in content that are 2+ years old
  2. Broken external links: links to pages that now 404
  3. Stale references: mentions of deprecated tools, old versions, discontinued products
  4. Competitor improvement: competitors have published newer, more comprehensive content on the same topic
  5. Thin content: pages with < 500 words that once ranked but may have been surpassed
  6. Missing current year: "best X" or "guide to X" content without current year reference

Detection method:

  1. Scrape the page content with firecrawl
  2. Look for year mentions older than 2024
  3. Check for "updated" or "published" dates > 12 months old
  4. Compare word count against current top 5 for the same keyword
  5. Check if competitors have fresher content via search dates

Output:

{
  "url": "/blog/old-post",
  "title": "Page Title",
  "decay_signals": ["outdated_statistics", "thin_content"],
  "published_date": "2023-06-15",
  "word_count": 450,
  "competitor_avg_word_count": 1800,
  "refresh_priority": "high",
  "recommended_actions": ["Update statistics to 2026", "Expand content to 1500+ words", "Add FAQ section"]
}

Pre-Publish Checklist

Generate a reusable markdown checklist:

## Pre-Publish SEO Checklist

### Target Keyword: _______________

#### Content
- [ ] Word count >= competitor average (target: ___ words)
- [ ] Primary keyword in first 100 words
- [ ] Keyword density 1-2%
- [ ] LSI/semantic keyword variations used
- [ ] Direct answer to target query in first paragraph (for AEO)
- [ ] Readability grade 7-9 (Flesch-Kincaid)

#### HTML Tags
- [ ] Title tag: 50-60 chars, keyword in first half
- [ ] Meta description: 120-155 chars, includes CTA
- [ ] Exactly one H1 containing primary keyword
- [ ] H2/H3 hierarchy correct, keywords in subheadings
- [ ] URL slug short, contains keyword, no parameters

#### Media
- [ ] All images have descriptive alt text
- [ ] Images in WebP/AVIF format
- [ ] Images have width/height attributes
- [ ] Below-fold images have loading="lazy"

#### Links
- [ ] Min 3 internal links with relevant anchor text
- [ ] Links to pillar/cluster content
- [ ] Min 1 external link to authoritative source
- [ ] No broken links

#### Schema & Social
- [ ] Appropriate JSON-LD schema added
- [ ] Open Graph tags (og:title, og:description, og:image)
- [ ] Twitter Card tags

#### E-E-A-T
- [ ] Author byline with credentials
- [ ] Sources cited for claims/statistics
- [ ] Published date and "last updated" date

#### AEO (AI Search)
- [ ] Direct answer in first 50 words
- [ ] FAQ section with FAQPage schema
- [ ] Content structured with clear headings for AI extraction

Title/Meta Description Generator

Given a target keyword and page content, generate optimized options:

Title tag rules:

  • 50-60 characters
  • Primary keyword in the first half
  • Include power word (Best, Guide, How to, Complete, Free, Proven)
  • Include current year for time-sensitive content
  • Brand name at end if space allows: | Brand

Meta description rules:

  • 120-155 characters
  • Include primary keyword naturally
  • Include a CTA (Learn, Discover, Find out, Get started, Compare)
  • Create urgency or curiosity when appropriate
  • Unique per page — never duplicate

Generate 3-5 title options and 3-5 meta description options for the user to choose from.

平台分布

Codex

34.86%
按下载量换算25

Claude

30.91%
按下载量换算23

Cursor

18.25%
按下载量换算13

Gemini CLI

9.47%
按下载量换算7

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

可疑

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。来源安全扫描存在提醒或未通过项,安装前请自行复核。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