AlphaMoatによると、2026-10時点でSEO (Site Audit + Content Writer + Competitor Analysis)のダウンロード数は26,927、スター数は34で、累計ダウンロード数で今回のランキング対象AIスキル10,914件中1,046位、コンテンツ制作では927件中116位です。
一覧の初期表示は累計ダウンロード数順で、スター数順にも並べ替えできます。詳細ページの順位は累計ダウンロード数で計算しています。
| プラットフォーム | ダウンロード数 | スター数 | コメント |
|---|---|---|---|
| ClawHub ↗ | 1.9万 | 31 | - |
| SkillHub ↗ | 2.7万 | 34 | — |
プラットフォームごとに集計方法と集計時点が異なります。
インデックスと技術面の健全性、ページ内の改善、内容とリンクの構成を扱い、手動の措置、インデックス、意図の一致、技術、内容、リンクの順で診断します。順位や自然流入の下落、ページのインデックス除外に対応し、対象の検索語に向けて内容を書き直したり統合したりできます。有料検索広告とアプリストアの最適化は対象外と明記し、リンクは自然に得るものだけを扱います。 SEO profile、keyword tracking、audit history、draft は ~/Clawic/data/seo/ に保存されます。再利用したくない場合は内容を確認または削除してください。提案を公開・実装する前に、最新の Search Console、実際の検索結果、対象 URL に照らして検証します。
資料公開日 2026-10-27
スキルのドキュメントをもとに生成した例であり、実際の実行結果ではありません。
2026-09 から集計方法を変更し、同名スキルを作者ごとに分けて集計し、全スキルを対象にしました。破線は変更前の月で、変更後のデータとは直接比較せず、変更をまたぐ増減も計算していません。 同名の別スキルがあるため、変更前のデータは現在ダウンロード数が最も多いこのスキルに帰属させています。
| 月 | ダウンロード数 | 前月比 | スター数 |
|---|---|---|---|
| 2026-10 | 2.7万 | — | 34 |
| 2026-09 | 1.8万 | — | 31 |
| 2026-08 | 1.8万 | +4.2% | 35 |
| 2026-07 | 1.7万 | +9.1% | 34 |
| 2026-06 | 1.6万 | +5.2% | 33 |
| 2026-05 | 1.5万 | — | 28 |
| 2026-03 | 1.1万 | — | 25 |
作者が公開した元のファイルをテキストのまま表示しています。AlphaMoat による編集や審査はしていません。リンクはクリックできません。コマンドは内容を確認してから実行してください。取得日 2026-10-23。
---
name: seo
slug: seo
version: 1.0.6
description: "Improves Google rankings with search engine optimization (SEO): audits, keyword research, content optimization, technical fixes, schema, and links. Not for paid search ads (PPC/SEM) or app store optimization. Use when organic traffic or rankings drop, pages are not indexed or get deindexed, crawl errors appear in Search Console, or a migration, redesign, or domain change is planned. Also for writing or refreshing content to rank, picking target keywords, fixing titles, canonicals, sitemaps, robots.txt, redirects, hreflang, duplicate content, or Core Web Vitals, and adding schema markup. And for chasing featured snippets or AI Overview citations, recovering from a Google penalty, core update, or manual action, optimizing local, ecommerce, SaaS, or publisher sites, or setting up Bing Webmaster Tools and IndexNow."
homepage: https://clawic.com/skills/seo
changelog: "Full coverage pass: deeper guides, situation-named files, and per-user configuration"
metadata:
clawdbot:
emoji: 🔍 os:
- linux
- darwin
- win32
displayName: SEO (Site Audit + Content Writer + Competitor Analysis)
configPaths:
- ~/Clawic/data/seo/
---
All persistent data (site profiles, config, audit history, keyword tracking) lives in `~/Clawic/data/seo/`. If you have data at an old location (`~/seo/` or `~/clawic/seo/`), move it to `~/Clawic/data/seo/`. Read `setup.md` on first use; `memory-template.md` holds the file format.
## When To Use
- Auditing a site: indexing, technical health, on-page, content, links, structure
- Diagnosing a ranking or organic traffic drop, a deindexed page, or a core-update hit
- Writing, refreshing, or consolidating content to rank for a target query
- Keyword research, opportunity sizing, competitor gap analysis, cannibalization
- Shipping structural changes safely: migrations, redesigns, hreflang, faceted navigation, programmatic pages
- Earning visibility in the surfaces above the ten blue links: snippets, packs, AI Overviews
- Not for paid search ads (PPC, Google Ads bidding, quality score) or for content planning with no search target — route to `content-marketing`
## Quick Reference
| Situation | Read |
|-----------|------|
| Ranking or organic traffic dropped | Triage below first, then `recovery.md` |
| Core update hit, manual action, hacked or spammy pages, reconsideration request | `recovery.md` |
| First pass on a new site, scoping an audit, writing the deliverable | `audits.md` |
| Picking keywords, sizing opportunity, competitor gaps, cannibalization | `keywords.md` |
| Deciding what to write, intent, E-E-A-T, thin content, AI-written drafts | `content.md` |
| Bottom-funnel pages: alternatives, "vs", best-X, pricing, integrations, glossary | `commercial-pages.md` |
| Titles, meta descriptions, headings, URLs, image alt text | `on-page.md` |
| Site structure, click depth, topic clusters, subdomain vs subfolder, index bloat, new site from zero | `architecture.md` |
| Pages not indexed, robots.txt, canonicals, sitemaps, status codes, duplicates, log files | `technical.md` |
| Slow pages, LCP / INP / CLS failing, PageSpeed vs field data | `performance.md` |
| React/Vue/Angular site, content missing from view-source, SPA routing, soft 404s | `javascript.md` |
| JSON-LD, rich results, review stars, breadcrumbs, product feeds | `schema.md` |
| Internal links, backlinks, anchor text, outreach, disavow | `links.md` |
| Physical location or service area, Google Business Profile, reviews, map pack | `local.md` |
| Product and category pages, faceted navigation, variants, out-of-stock, pagination | `ecommerce.md` |
| Multiple languages or countries, hreflang, geotargeting, translated duplicates | `international.md` |
| Domain change, replatform, redesign, URL restructure, HTTP to HTTPS | `migrations.md` |
| Featured snippets, People Also Ask, sitelinks, image and video results, pixel position | `serp-features.md` |
| AI Overviews, AI Mode, ChatGPT and Perplexity citations, AI crawlers, llms.txt | `ai-search.md` |
| Search Console data, exports, regex filters, proving a change worked, forecasting, reporting | `search-console.md` |
| Bing, IndexNow, Applebot, Yandex, Naver, Baidu, privacy engines | `other-engines.md` |
| Publisher work: Top Stories, Google News, Discover traffic, paywalls | `news-discover.md` |
| Generating hundreds or thousands of templated pages | `programmatic.md` |
| WordPress, Shopify, Webflow, Wix, Next.js platform quirks | `cms-platforms.md` |
| Anything else SEO | Run the audit checklist below, then route from the area that fails |
## Core Rules
1. **Audit before prescribing.** Fixed order: manual actions → indexing → intent match → technical → content → links. Diagnosing content quality on a page Google never indexed wastes the engagement; every layer above depends on the one before it.
2. **The SERP is the spec.** Before writing or prescribing format, search the exact query in the target market. Page 1 defines format, depth, and freshness. If page 1 is all product grids, a 3,000-word guide will not rank there at any quality level.
3. **Improve before you create.** Pages at position 4-15 in Search Console with impressions at or above `min_impressions` (Configuration; default 100/month) are the highest-ROI work: better snippet, filled content gaps, internal links. Write a new page only when no existing URL targets the intent. Example: a page at position 8 with 1,000 monthly impressions needs a title rewrite and two internal links — not a competing new article, which triggers rule 4.
4. **One intent per page.** Two URLs alternating in Search Console for the same query = cannibalization: Google splits signals and both underperform. Fix: 301 the weaker into the stronger, merge unique content. Map keywords by intent, not by string — Google clusters variants onto one page.
5. **Technical floor.** Core Web Vitals "good" thresholds (Google, field data at p75): LCP < 2.5s, INP < 200ms, CLS < 0.1. HTTPS, mobile-first, self-referencing canonicals, clean sitemap. Technical debt caps everything content can earn.
6. **E-E-A-T is signals, not a score.** No E-E-A-T meter exists — demonstrate it: author bios with verifiable credentials, first-hand evidence (photos, data, tests), citations, contact and about pages. Decisive for YMYL topics (health, finance, legal, safety).
7. **Links: earn, never buy.** Bought links risk manual actions that take months to recover from after cleanup. Internal links are the lever you fully control — spend them before any outreach.
8. **Iterate from Search Console, not assumptions.** Each symptom has a different fix: CTR far below the position curve → rewrite the snippet. Impressions but no clicks → intent or SERP-feature problem. No impressions → indexing or relevance problem. GSC tells you which; guessing does not.
9. **Size the prize before doing the work.** `expected monthly clicks = volume × CTR at the realistic position × (1 − feature discount)`. Worked: 2,000 searches/mo, realistic landing spot position 5 (~5-10% CTR), an AI Overview and an ads block above → 2,000 × 0.07 × 0.6 ≈ 84 clicks/mo. If that number does not justify the build, say so before writing.
## Ranking Drop Triage
In order — stop at the first confirmed cause:
1. GSC → Manual Actions + Security Issues. If flagged, that IS the diagnosis (`recovery.md`).
2. URL Inspection on hit pages: still indexed? Google-selected canonical changed to another URL?
3. Recent deploys: robots.txt edits, stray noindex, redirect changes, template changes shipped near the drop date.
4. Drop date vs Google's announced update dates. Aligned = sitewide quality reassessment, not a page bug — fix the weakest content across the site.
5. Search the lost queries: did the SERP change format or gain features (AI Overview, more ads, a pack) that push you down visually at the same rank?
6. Backlinks: links lost on your side, or a competitor gained them.
7. Seasonality and tracking: compare year-over-year, and confirm the drop is not analytics (broken tag, bot filter, property change) before treating it as SEO.
8. None conclusive → gap-compare the pages that replaced you (`keywords.md`).
## SEO Audit Checklist
**Indexing:**
- [ ] Important pages indexed — verify with URL Inspection and the Page indexing report, not the `site:` operator
- [ ] No important pages blocked in robots.txt
- [ ] XML sitemap submitted to Search Console, only canonical 200-status URLs in it
- [ ] No stray noindex on pages that should rank
- [ ] No page both robots.txt-blocked AND noindexed — blocked crawl means Google never sees the noindex, so the page can stay indexed
**Technical:**
- [ ] Core Web Vitals passing (thresholds in rule 5)
- [ ] Mobile-friendly, HTTPS with no mixed content
- [ ] No crawl errors, no soft 404s, no server errors in Search Console
- [ ] Redirect chains ≤3 hops
- [ ] Rendered HTML contains the main content (JavaScript sites)
**On-Page:**
- [ ] Unique title tags (50-60 chars), meta descriptions (150-160 chars)
- [ ] One H1 per page with the target term; proper heading hierarchy
- [ ] Images with alt text; internal links to and from the page
**Content:**
- [ ] Search intent matched (rule 2)
- [ ] No cannibalization (rule 4)
- [ ] No thin or duplicate content; dead pages improved, consolidated, or removed
**Off-Page and Entity:**
- [ ] Google Business Profile complete (local businesses)
- [ ] Backlink profile checked for toxic patterns
- [ ] Brand queries return the right result, sitelinks, and knowledge panel where applicable
## Content Writing Process
1. **Keyword research** — target keyword, volume, SERP reality check (`keywords.md`)
2. **Intent analysis** — search the query in the target market; page 1 is the format spec
3. **Gap + gain** — cover what ranking pages cover, then add what none of them have
4. **Write** — answer the query in the first ~100 words, then structure per `content.md`
5. **Optimize** — title, meta, headers, internal links from strong pages, schema
6. **Publish** — request indexing once, log in `~/Clawic/data/seo/memory.md`, review GSC after a few weeks
## What Takes How Long
Honest ranges beat invented dates; every one of these is mechanism, not promise.
| Change | Time to effect | Mechanism |
|---|---|---|
| Title or meta rewrite | Days to appear, 2-4 weeks to judge | Needs a recrawl, then enough impressions for CTR to be readable |
| New page, established site | Indexed in days; competitive movement 3-6 months | Discovery is fast; earning position takes links and engagement history |
| New domain | Months before competitive queries move | No confirmed sandbox — the delay is missing links, history, and coverage |
| Core Web Vitals fix | ~4 weeks before field data reflects it | CrUX field data is a 28-day rolling window; lab scores change instantly |
| 301 migration | 2-8 weeks of turbulence, longer on large sites | Google must recrawl every redirected URL to transfer signals |
| Manual action revoked | Days to weeks after reconsideration | Human review queue; recovery of position is separate and slower |
| Core update recovery | Usually at the next update, not between them | Sitewide reassessments are recomputed on update cycles |
| Disavow file | Weeks to months | Applied as links are recrawled, not on upload |
## Output Gates
Before delivering recommendations or content:
- Did I search the actual query in the target market, or am I prescribing format from memory?
- Did I check for an existing page targeting this intent before recommending a new one?
- Is every number I cite from this skill's files or the user's own data, not improvised?
- Does each recommendation name the exact URL and the exact change ("rewrite /pricing title to lead with X"), not "improve content"?
- Did I size the prize (rule 9) so the user can refuse the work?
- Am I promising a date? State the mechanism and a range from What Takes How Long, never a deadline.
## Configuration
User-dependent variables. Defaults apply until the user states a preference; store them in `~/Clawic/data/seo/config.yaml`.
| Variable | Type | Default | Effect |
|---|---|---|---|
| site_type | blog \| ecommerce \| saas \| local \| news \| directory \| auto | auto | Weights the audit checklist and decides which guide the router opens first; `auto` infers from URL patterns and page templates on first look |
| target_market | text (locale, e.g. en-US) | en-US | Which SERP to check, which spelling variant to write, and whether hreflang and local guidance apply |
| tool_access | gsc-only \| paid-suite | gsc-only | `gsc-only` keeps every workflow on free data (GSC, Trends, SERPs); `paid-suite` unlocks backlink-index and difficulty-score steps |
| risk_posture | conservative \| standard \| aggressive | conservative | Gates link tactics, external anchor ratios, and how much templated page generation to recommend |
| cms | wordpress \| shopify \| webflow \| wix \| headless \| other \| auto | auto | Picks the implementation path for every fix (where redirects, robots, and metadata actually live) |
| min_impressions | number (impressions/month) | 100 | Floor for every opportunity list: striking distance (rule 3, `keywords.md`, `search-console.md`) and the traffic-at-stake cutoff for the audit Top 5. Below it, a page cannot produce a readable CTR or click change — raise it on large sites |
| voice_file | path | none | Brand voice guide at `~/Clawic/data/seo/<file>`; governs drafted copy, never the SEO structure |
Preference areas — customizable dimensions; a stated preference gets recorded in config.yaml and applied:
- **Reporting**: audit depth (one-page priorities vs full report), technical vs business language, which KPI leads (clicks, conversions, revenue)
- **Conventions**: URL and slug style, title formula (brand position and separator), heading patterns, internal anchor style
- **Scope boundaries**: sections that are off-limits (legal, careers, docs), staging hosts, URLs never to touch
- **Implementation**: whether the agent edits files and writes patches or hands over specs and tickets
- **Measurement**: reporting property, comparison window, whether branded queries are excluded from "SEO traffic"
- **Thresholds**: floors that gate what gets worked on — impressions (`min_impressions`), traffic at stake for a finding to make the Top 5, minimum sample before a test is called
- **Cadence**: rank and GSC review frequency, content refresh schedule, reporting date
## Traps
| Trap | Why it fails | Do instead |
|------|-------------|------------|
| Writing before checking the SERP | Format mismatch = no ranking at any quality | Rule 2: the SERP is the spec |
| New article for a query an existing page ranks 4-15 for | Cannibalization splits signals | Improve the existing page (rule 3) |
| Chasing keyword density | No density threshold exists; stuffing detection is pattern-based | Cover the topic, use variants naturally |
| noindex on a robots.txt-blocked page | Google never crawls it, never sees the noindex | Allow crawl until deindexed, then block |
| Changing URLs without 301s | Links and authority now point at 404s | Map every old URL to its closest new match |
| Buying links or PBNs | Payment and network footprints → manual action | Earn links via assets and digital PR |
| Reporting rank without checking the rendered SERP | #1 under an AI Overview and four ads earns a fraction of historical #1 clicks | Check pixel position, not just rank |
| Judging index coverage with `site:domain.com` | The operator is an estimate and excludes results Google chooses to hide | Page indexing report + URL Inspection |
| Re-requesting indexing for the same URL repeatedly | The queue is not a priority auction; nothing accelerates | Fix the reason it was not indexed (quality, duplicate, discovery) |
| Fixing every warning a crawler emits | Crawl tools flag non-signals (meta keywords, long titles on pages with no impressions) | Rank issues by the traffic at stake, then fix |
| Optimizing a page whose Google-selected canonical is another URL | Every signal you add credits the other URL | Resolve the canonical conflict first |
| Reporting "average position improved" as a win | Average position moves when the query mix changes; new long-tail impressions drag it down while traffic grows | Report clicks and conversions, positions per query |
## Where Experts Disagree
- **Word count.** Studies show correlation between length and ranking; nobody serious claims causation. Cover the intent fully, then stop — padding is a negative.
- **Disavow.** Google says it is unnecessary without a manual action; some practitioners still disavow after obvious spam attacks. Default: don't, unless a manual action names links.
- **Exact-match anchors.** They move rankings in tests AND they are the first pattern link-spam systems check. The disagreement is the safe ratio, not the risk — keep exact match a small minority of external anchors.
- **Subdomain vs subfolder.** Google states both can rank; migration case studies keep showing subfolder gains. Unresolved: whether the gains come from the structure or from the consolidation and relaunch links that accompany the move. Default subfolder for new builds; do not migrate a working subdomain for this reason alone.
- **Optimizing for AI Overviews.** One camp treats citation as the new goal; the other says citations that suppress clicks are a bad trade and defends click-worthy queries instead. Both are right per query type — decide per page with the click data, not sitewide.
## Related Skills
More Clawic skills, get them at https://clawic.com/skills/seo (install if the user confirms):
- `content-marketing` — Content strategy
- `analytics` — Traffic analysis
- `market-research` — Competitive analysis
- `html` — HTML optimization
- `web` — Web development
## Feedback
- If useful, star it: https://clawic.com/skills/seo
- Latest version: https://clawic.com/skills/seo
Part of [Clawic](https://clawic.com), the verified skill library. Get this skill: https://clawic.com/skills/seo.
李白.Skill:润色专家 检测并消除中文文本中的AI生成痕迹,将生硬的机器输出转化为自然流畅的人类文笔。融合规则检测、分层改写和质控流水线,确保最终文本通过人工审查。
李白.Skill:润色专家 检测并消除中文文本中的AI生成痕迹,将生硬的机器输出转化为自然流畅的人类文笔。融合规则检测、分层改写和质控流水线,确保最终文本通过人工审查。
文章に残る AI らしさを見つけて取り除き、人が書いたように読める文章にします。
【零一数科·出品】视频号爆款短视频拆解(全能版)(付费版)。视频号爆款短视频拆解(付费版:全能)。一键拆解爆款,把一条视频号视频拆成结构分段、爆款归因、六维评分和可借鉴策略,照着学照着抄。支持视频号分享链接和本地视频,付费全能版远端分析,1–5 分钟出可视化 HTML 报告。
【零一数科·出品】视频号挂车爆单短视频脚本生成(付费版)。一条链接二创脚本,丢进爆款视频号链接即以此骨架改写成你的挂车脚本,或按方向/目的/商品原创多条口播脚本(含分段表、口播全文、六维质检评分)。
AI が書いた文章の典型的な痕跡を見つけ、意味を保ったまま自然な文章に書き直す文章編集スキルです。
用于中文公文、事务性材料和新闻稿件的起草、改写、压缩、润色、审校、文种核对、去口语化、降 AI 味及 Word 格式处理,适用于机关、企事业单位、学校和新闻机构。涵盖申请、请示、报告、通知、通告、意见、决定、决议、议案、公报、命令、函、复函、批复、说明、方案、纪要、公告、公示、通报、制度、规定、办法、细则、操作规程、工作要点、总结、调研、讲话、致辞、主持词、述职、可研、审查材料、技术需求、新闻消息、编者按、新闻评论,以及采购、整改、反馈和 AI 算力等场景。
【零一数科·出品】视频号爆款文案生成(体验版)。免费免登录出脚本,给创作目的、行业、受众、选题就按爆款结构(Hook-中段-CTA、节奏标签)产出口播脚本,信息不全也能按爆款潜质兜底。本地运行、无需 Key,零成本上手。
利用者の訂正と自己点検を層構造の記憶に蓄え、使われ方に応じて昇格や保管を行うスキルです。
数式、書式、ブックの構造を保ったまま Excel ファイルを作成・点検・編集するスキルです。
Word と DOCX 文書を扱うための作業指針で、編集後にスタイルや番号、レイアウトが崩れないことに重点を置きます。
複数の動画モデル API をまたぐ AI 動画生成と編成のスキルです。
データベースや表計算、書き出した生データを意思決定に使える結論へ変えるデータ分析スキルです。
データ月:2026-10 · ダウンロード数とスター数は公開スキルレジストリ(ClawHub、SkillHub)から月単位で集計しています。詳しくはデータの方法論をご覧ください。