
SEO-IntentRank (Antigravity MCP)
io.github.AI-BuildInfra由 GitHub 帳號 AI-BuildInfra 發布v1.0.1更新於 Oct 24, 2026
Human search intent keywords, Core Web Vitals, technical SEO & on-page audit MCP server.
發布方自己的服務:Rank Fast
概覽
本機 MCP 伺服器,用於網頁 SEO 稽核:搜尋意圖關鍵字、頁面標籤、Core Web Vitals 與技術結構。
- 可以做什麼
- 從網址、HTML 或純文字擷取人類搜尋意圖關鍵字、意圖分布、自然搜尋問句與已解答問題
- 稽核頁面 SEO:標題長度與像素寬度、meta 描述、Open Graph、Twitter Card、canonical 以及圖片 alt 與尺寸檢查
- 透過 Google PageSpeed Insights 取得 Lighthouse Core Web Vitals(CLS、LCP、INP、FCP、TTFB),離線時改用本機 DOM 啟發式分析
- 檢查技術 SEO:標題層級、Schema.org JSON-LD、robots.txt 與 sitemap.xml、文字與 HTML 比例
- 執行綜合稽核,輸出 0-100 分數、字母等級與依優先順序排列的改善建議
- 適用情境
- 適合讓助理檢查某個頁面或網站的 SEO 與效能問題並提出修正建議,例如發布內容前自我檢查,或排查版面位移、搜尋意圖涵蓋不足等狀況。支援線上網址,也支援貼上 HTML 離線分析。
- 執行需求
- 以 stdio 方式在本機執行,透過 npx 或 node 啟動,需要 Node.js。未宣告帳號或環境變數。稽核線上網址需要網路連線;Lighthouse 工具可選擇以工具參數傳入 Google PageSpeed Insights API 金鑰。
安裝
在 SourceWeft 中
- 開啟 儀表板中的 SEO-IntentRank (Antigravity MCP),將其新增到工作區。
- 為需要使用其工具的對話啟用該服務。
Desktop only,透過 STDIO。 STDIO 服務會啟動本機處理程序,因此需要 SourceWeft 桌面主機。
其他 MCP 客戶端
參照 儲存庫 中的啟動說明。
README
SEO-IntentRank MCP Server (seo-intentrank)
[AI Build Infra Website] [Maintained by AI Build Infra] [Model Context Protocol] [M8ven Score] [CI Status] [NPM Version] [MIT License]
Official Model Context Protocol (MCP) Server for SEO-IntentRank: Human Search Intent Keywords, Technical SEO Audits, Google Lighthouse Core Web Vitals (CLS/LCP/INP), and On-Page Diagnostic Intelligence.
Developed, maintained, and published by AI Build Infra — High-Performance Infrastructure for Artificial Intelligence & Search Engine Optimization.
🌟 Overview
SEO-IntentRank is an enterprise-grade SEO diagnostic MCP Server created by AI Build Infra to equip AI assistants (ChatGPT, Claude Desktop, Google Antigravity, Cursor, and Perplexity) with live web diagnostics.
It bridges the gap between raw technical crawl data and real human search behavior, delivering deep on-page audits, direct Google Lighthouse Core Web Vitals (including exact Cumulative Layout Shift / CLS scores), and semantic Human Search Intent Keyword extraction.
Core Capabilities:
- 🎯 Human Intent Keyword Engine: Categorizes page search intent into Informational, Commercial Investigation, Transactional, and Navigational; synthesizes high-converting natural human search queries; mines answered questions; and flags AI slop clichés and keyword stuffing.
- 🔍 On-Page SEO Inspection: Scans
<title>(character length & ~580px pixel width estimates),<meta name="description">(CTA detection),<meta name="robots">, Open Graph tags (og:*), Twitter Cards (twitter:*), and Canonical URL verification. - 📐 Media & Image Alt Audit: Detects missing
altattributes, decorative images (alt=""), and missingwidth/heightattributes that cause Cumulative Layout Shift (CLS). - ⚡ Lighthouse Core Web Vitals: Fetches exact numeric scores directly from Google Lighthouse / PageSpeed Insights for:
- Cumulative Layout Shift (CLS) with threshold analysis ($\le 0.1$ Good) and shift element culprit identification.
- Largest Contentful Paint (LCP), Interaction to Next Paint (INP), First Contentful Paint (FCP), and Time to First Byte (TTFB).
- Fallback to local DOM-based layout shift heuristics when auditing raw HTML offline.
- 🏗️ Technical SEO Hierarchy: Checks single
<h1>enforcement, non-sequential heading level skips (e.g. H1 → H3), Schema.org JSON-LD structured data validation,robots.txtdisallows,sitemap.xml, and text-to-HTML ratios.
💡 How SEO-IntentRank Works in AI Assistants (ChatGPT / Claude / Antigravity)
🚀 Installation & Quick Start
Running with NPX
Adding to Antigravity / Claude Desktop / Cursor Configuration
Add the following to your mcpServers configuration (mcp_config.json or claude_desktop_config.json):
Or for local development:
🛠️ Available MCP Tools
1. extract_human_intent_keywords
Extracts high-converting Human Search Intent Keywords and analyzes audience search behavior.
Parameters:
url(string, optional): Live URL.html(string, optional): Raw HTML markup.text(string, optional): Plain article text.targetKeyword(string, optional): Target focus keyword to evaluate.
Output:
- Primary search intent & percentage distribution (
informational,commercial,transactional,navigational). - Top ranked human intent keywords with relevance scores and heading signals.
- Synthesized natural human search queries answered by the page.
- Mined questions with answered status.
- Information Gain & AI slop fluff assessment score.
2. scan_onpage_seo
Audits On-Page SEO elements from a live URL or raw HTML string.
Parameters:
url(string, optional): Web URL to fetch and audit.html(string, optional): Raw HTML string for offline analysis.
Output:
- Title status, length, and pixel width estimate (~580px Google cutoff).
- Meta description length & CTA presence.
- Complete Open Graph & Twitter Card validation.
- Canonical tag status (self-referencing, missing, relative, or mismatched).
- Image Alt tag audit with missing dimensions (CLS risk detection).
3. audit_lighthouse_cls_vitals
Audits Core Web Vitals directly using Google Lighthouse / PageSpeed API and local DOM shift heuristics.
Parameters:
url(string, optional): Live URL to benchmark.html(string, optional): HTML snippet for local layout shift risk calculation.strategy(string, optional):'mobile'(default) or'desktop'.apiKey(string, optional): Optional Google PageSpeed Insights API key.
Output:
- Cumulative Layout Shift (CLS) score, rating (
GOOD,NEEDS_IMPROVEMENT,POOR), and layout shift culprit elements. - LCP, INP, FCP, TTFB measurements.
- Lighthouse category scores (Performance, SEO, Accessibility, Best Practices).
4. audit_technical_seo
Inspects technical page architecture and semantic structure.
Parameters:
url(string, optional): Live URL.html(string, optional): Raw HTML string.robotsTxtContent(string, optional): Rawrobots.txtstring to parse.sitemapXmlContent(string, optional): Rawsitemap.xmlstring to parse.
Output:
- Heading structure validation (H1 count, non-sequential heading level skips).
- Schema.org JSON-LD syntax and
@typeverification. - Content word count, text-to-HTML ratio, and thin content warnings.
- Robots directives and Sitemap URL counts.
5. comprehensive_seo_audit
Master all-in-one SEO auditor running all tools in parallel and producing an executive report.
Parameters:
url(string, optional): Live URL.html(string, optional): Raw HTML markup.strategy(string, optional):'mobile'|'desktop'.apiKey(string, optional): Optional Google PageSpeed API key.targetKeyword(string, optional): Target focus keyword.
Output:
- Overall SEO score (0-100) and letter grade (
A+,A,B,C,D,F). - Aggregated On-Page, Technical, Lighthouse, and Intent audits.
- Prioritized action plan sorted by impact (
HIGH,MEDIUM,LOW).
🧪 Testing
Run the automated test suite:
🏢 Publisher & Entity Reconciliation
SEO-IntentRank is engineered and published by AI Build Infra.
- Official Website: https://aibuildinfra.com/
- Publisher: AI Build Infra
- Ecosystem: High-performance developer tools, E-E-A-T entity reconciliation, and AI agent infrastructure.
- Related MCP Servers: HumanCraft UI (
@aibuildinfra/humancraft)
📄 License
MIT License. Created with ❤️ by AI Build Infra.
來源:README.md,提交 77e3f36
工具
0版本歷史
1- v1.0.1最新Sep 24, 2026
評分
已滿足適用檢查項的 50%
該評分衡量此收錄條目的完整程度,依據註冊中心條目、端點與公開來源計算得出。它不是對伺服器本身的評價,也不表示伺服器安全或目前可用:可連通性是帶日期的檢查結果,評分也不包含安裝量或使用者評價。
- 近 30 天內可連通25 分
請保持端點能夠回應 MCP 握手;SourceWeft 每 30 天探測一次。
- 已宣告認證方式10 分
請在註冊中心條目中宣告憑證請求標頭,或保持端點能夠回應探測,以便判定其認證方式。
- 已列出工具15 分
請讓伺服器在無憑證時回應 tools/list,或發布可在沙箱中無憑證啟動的軟體包。
- 描述必需15 分
- 圖示10 分
- 文件15 分
- 授權條款10 分
其他 Rank 整合
Rank Fast
fast.rank官方發布者
Rank Fast 遠端 MCP 伺服器,讓助理檢查 Google 收錄並提供診斷、取得自動完成關鍵字建議,並檢視 AI 能見度與客座文章資源。

dep-rank
由 GitHub 帳號 j7an 發布
依星標數或信任啟發式為依賴某個 GitHub 儲存庫的專案排名,並可在這些依賴方中搜尋程式碼。

AIWiseRank
由 GitHub 帳號 VLCODE87 發布
一個遠端 MCP 伺服器,讓助理依真實網路流量搜尋、排名與比較 AI 工具,並提供價格、免費方案、替代品和 LLM API 基準資料。

Mcp Local Rank Map Pack Tracker
com.mambabuilt
本機 MCP 伺服器,透過 Apify actor 查詢指定地點下關鍵字的 Google 自然排名、地圖包排名與本地搜尋器排名。
CreceRank
com.crecerank
遠端 MCP 伺服器,用來檢視品牌在 ChatGPT、Gemini、Perplexity 和 AI Overviews 等 AI 助理中的曝光情形,包含提及、來源與 GEO 建議。