Communitygithub.com

cygnixlabs-cmd/cygnix-seo-skills

为单个网页提供0-100的AI搜索可见性（GEO/AEO就绪度）评分，基于其服务端渲染的HTML——评估AI爬虫访问、Organization/WebSite JSON-LD、结构化数据、标题/元数据、问句式标题、答案优先开头、llms.txt及新鲜度——并给出前5个修复建议。当用户询问页面是否已为ChatGPT、Perplexity、Google AI Overviews或其他AI搜索做好准备，或希望获得GEO/AEO评分时使用。

cygnix-seo-skills 是什么？

这是一个针对单个网页的AI搜索可见性快速评分技能，核心价值在于评估页面在AI爬虫眼中的就绪度。它直接抓取服务端渲染的HTML（不执行JavaScript），模拟大多数AI爬虫的实际行为，给出0-100的分数及对应的等级（STRONG/GOOD/FAIR/POOR/CRITICAL）。技能覆盖8个关键检查项：AI爬虫在robots.txt中的访问权限、Organization/WebSite等实体JSON-LD、其他有效JSON-LD、标题/元描述长度、问句式标题结构、答案优先开篇及内容字数、llms.txt文件存在性、以及日期信号。每次运行后，它会返回最值得优先修复的5个问题。该技能适用于ChatGPT、Perplexity、Google AI Overviews等AI搜索引擎，也适用于任何想要评估GEO（生成式引擎优化）或AEO（答案引擎优化）就绪度的用户。注意：该评分只衡量爬虫的可抓取性和可信度，而非当前的实际引用情况。JS渲染为主的网站得分会偏低，这反映的是AI爬虫无法执行JavaScript的真实情况。

兼容平台~Claude Code~Codex CLI~Cursor
npx skills add https://github.com/cygnixlabs-cmd/cygnix-seo-skills/tree/HEAD/skills/ai-visibility-quick-score
查看原文→浏览所有技能

在你喜欢的 AI 中提问

打开一个已预加载此 Agent Skill 的新对话。

OpenAIChatGPTClaudeClaudeGeminiGeminiGrokGrokPerplexityPerplexityDeepSeekDeepSeek

文档

AI可见性快速评分（单页面）

根据大多数AI爬虫的实际抓取方式——原始HTML，不执行JavaScript——为单个URL评分。返回0–100分、一个等级以及最值得修复的5个问题。

何时使用

  • "我的首页为AI搜索做好准备了吗？"、"这个URL的GEO评分是多少？"、"为什么ChatGPT没有提到我们？"（作为初步检查）。

如何运行

python3 scripts/quick_score.py https://example.com/

需要出站HTTPS；共发起3次请求（页面、robots.txt、llms.txt）。

评分标准（100分）

检查项分值
AI爬虫在robots.txt中被允许（每个被屏蔽的机器人扣5分）25
JSON-LD中包含Organization / WebSite / Person实体15
其他有效的JSON-LD（Article, Product, FAQPage……）10
标题15–65字符，元描述70–170字符10
一个H1以及问句形式的H2/H315
答案优先的开头段落，且服务端渲染字数超过30015
存在 /llms.txt5
日期信号（dateModified、article时间、）5

等级：80+ 强，65+ 好，50+ 一般，35+ 差，低于35 严重。

诚实看待结果

该评分衡量的是就绪度（AI系统能否获取、解析并信任该页面），而非你当前是否被引用。纯JavaScript网站因为服务端HTML为空而得分较低，这个低分对于不运行JavaScript的爬虫来说是准确的。

示例

https://plausible.io/ → 72/100 好；修复建议：添加页面类型JSON-LD、直接答案的开头、/llms.txt、日期信号以及更短的标题。

想要整站评分和报告？

Individual skills in this repo

This repo contains 4 individual skills — each has its own dedicated page.

cygnixlabs-cmd/cygnix-seo-skills

Compare two Google Search Console page exports (an earlier period and a recent one) and flag pages that are LOST, DECLINING, in STRIKING_DISTANCE, GROWING or STABLE, ranked by clicks lost. Use when the user asks which pages lost traffic, about content decay, what to refresh or update first, or uploads two GSC page exports. Offline, no API keys.

cygnixlabs-cmd/cygnix-seo-skills

Find queries where two or more of your own pages split Google impressions, from a Google Search Console query+page export, and rank them HIGH/MEDIUM/LOW. Use when the user asks about keyword cannibalization, competing pages, pages fighting for the same keyword, or uploads a GSC export with query and page columns. Offline, no API keys.

cygnixlabs-cmd/cygnix-seo-skills

Group a keyword list into topic clusters with a pillar keyword, majority search intent and suggested page type, so one page targets each cluster. Use when the user pastes keywords or a keyword CSV and asks to cluster, group, organise or map keywords to pages, or plan a content hub. Offline, no API keys.

cygnixlabs-cmd/cygnix-seo-skills

Check whether a website has a valid /llms.txt (llmstxt.org format, working links, size) and whether robots.txt lets AI crawlers such as GPTBot, OAI-SearchBot, ClaudeBot, PerplexityBot and Google-Extended in. Use when the user asks about llms.txt, AI crawler access, blocking or allowing AI bots, or whether ChatGPT/Claude/Perplexity can read their site. Read-only.

相关技能

steipete/sag

ElevenLabs text-to-speech with mac-style say UX.

community

steipete/oracle

Oracle CLI second-model review/debug/refactor/design with selected files, dry-run token checks, API or browser engine.

community

steipete/peekaboo

Capture and automate macOS UI with the Peekaboo CLI.

community

mattpocock/design-an-interface

Generate multiple radically different interface designs for a module using parallel sub-agents. Use when user wants to design an API, explore interface options, compare module shapes, or mentions "design it twice".

community

mattpocock/prototype

Build a throwaway prototype to flesh out a design — a runnable terminal app for state/business-logic questions, or several radically different UI variations toggleable from one route.

community

mattpocock/grill-me

Interview the user relentlessly about a plan or design until reaching shared understanding, resolving each branch of the decision tree. Use when user wants to stress-test a plan, get grilled on their design, or mentions "grill me".

community