AI AGENT SKILLS

SEO & GEO 2026 — Rank Google + AI Search

一个面向 Research 场景的 Agent 技能。原始说明:You're publishing content but Google doesn't rank you. Or worse — AI search engines cite your competitors, not you. This is the dual-engine playbook: rank on...

SKILL.md

SKILL.md


name: gingiris-seo-geo
description: |
You're publishing content but Google doesn't rank you. Or worse — AI search engines cite your competitors, not you. This is the dual-engine playbook: rank on Google AND get cited by ChatGPT/Perplexity/Claude simultaneously.

What's inside:
• E-E-A-T writing voice system (sound authoritative to both Google and AI)
• Keyword funnel strategy (TOFU/MOFU/BOFU → landing page mapping)
• JSON-LD schema templates (FAQPage, HowTo, Article, Product)
• Comparison/alternatives page SOP (highest-converting SEO content type)
• GEO content patterns for AI citation + AI crawler configuration
• IndexNow setup for instant indexing

Built from: gingiris.tools 0 → indexed in 90 days. AFFiNE SEO-driven discovery to 60K stars. 150+ AI startup consultations.

Triggers: "SEO" | "SEO audit" | "site audit" | "technical SEO" | "content SEO" | "SaaS SEO" | "programmatic SEO" | "SEO daily report" | "page-1 keywords" | "首页关键词" | "GEO" | "Generative Engine Optimization" | "answer engine optimization" | "AEO" | "AI search optimization" | "ChatGPT SEO" | "Perplexity SEO" | "Claude SEO" | "Google AI Overview" | "AIO" | "AI citations" | "AI 引用" | "AI referral traffic" | "AI 引流量" | "track ChatGPT traffic" | "GA4 AI sources" | "GSC" | "Google Search Console" | "GA4 organic" | "rank tracking" | "排名追踪" | "keyword rank" | "schema markup" | "Schema.org" | "JSON-LD" | "FAQPage schema" | "HowTo schema" | "structured data" | "rich results" | "featured snippet" | "IndexNow" | "E-E-A-T" | "keyword strategy" | "keyword funnel" | "keyword research" | "TOFU MOFU BOFU" | "comparison page" | "alternatives page" | "competitor comparison" | "internal links" | "internal linking" | "内链" | "orphan pages" | "link equity" | "backlinks" | "backlink audit" | "link building" | "外链" | "robots.txt" | "AI crawler" | "GPTBot" | "ClaudeBot" | "PerplexityBot" | "CTR optimization" | "title rewrite" | "hreflang" | "i18n SEO" | "CTA conversion" | "Convert Block" | "搜索优化" | "AI 搜索" | "内容 SEO" | "GEO 优化" | "生成式引擎优化" | "SEO 审计" | "网站审计"


SEO & GEO 双引擎增长手册

Language / 语言: 中文 | English


中文版

作者:Iris (生姜iris) | 版本:v1.0 (2026年4月)

核心理念

SEO 和 GEO 不是两件事,是一件事的两面。结构化数据同时服务传统搜索和 AI 搜索。

传统 SEO 解决「被搜到」,GEO 解决「被 AI 引用」。

2026 年的搜索流量已经分裂为两个入口:Google/Bing 的传统排名,和 ChatGPT/Perplexity/Claude 的 AI 回答。你的内容必须同时在两个战场上赢。

SEO + GEO 双引擎对比

| 维度 | SEO(传统搜索) | GEO(AI 搜索) |
|------|----------------|----------------|
| 目标 | Google/Bing 排名 | AI 回答中被引用 |
| 核心信号 | 反向链接 + 关键词匹配 | 结构化数据 + E-E-A-T 真实性 |
| 内容格式 | H2/H3 层级 + 长尾词覆盖 | 直接回答 + 对比表格 + FAQ Schema |
| 技术手段 | XML Sitemap + Core Web Vitals | IndexNow + robots.txt 开放 AI 爬虫 |
| 衡量方式 | GA4 自然流量 + Ahrefs 排名 | 手动检查 AI 引用频率 |

五个核心原则

  1. 从 BOFU 往上做 — 先做高意向关键词(定价、对比),再做教育型内容。转化率最高的词优先。
  2. 真实声音是最好的 E-E-A-T — 不要让 AI 写得像 AI。用创始人声音、真实数字、亲历故事建立可信度。
  3. 结构化 = 可引用 — Key Stats 表格、FAQ Schema、对比矩阵让 AI 引擎直接抓取引用。
  4. IndexNow 是 GEO 的基础设施 — 内容更新后秒级推送到 Bing/AI 搜索,不等爬虫。
  5. 对比页是 SEO 金矿 — 每个竞品一个独立页面,截获决策期用户。

详细指南

| 主题 | 文件 | 说明 |
|:---|:---|:---|
| 关键词漏斗策略 | references/keyword-strategy.md | TOFU/MOFU/BOFU 关键词矩阵与优先级 |
| SEO 技术基础 | references/seo-foundations.md | 核心页面优化、Schema、站点架构 |
| GEO 优化实操 | references/geo-optimization.md | IndexNow、AI 爬虫(搜索 vs 训练分流)、Schema 类型速查矩阵、AI 引用监测 |
| 内容生产 SOP | references/content-sop.md | 从选题到发布的完整流程 |
| E-E-A-T 写作声音 | references/writing-voice.md | 创始人声音系统,提升真实性信号 |
| 竞品对比页 SOP | references/comparison-pages.md | 页面结构、Schema、关键词嵌入 |
| 测量层 — SEO/GEO 数据系统 🆕 | references/measurement-layer.md | 四核心指标、Article Database、GSC API、AI 引流量、归因陷阱 |
| SEO 审计 SOP 🆕 | references/seo-audit.md | 四相位审计(Technical→On-Page→Content→Off-Page)+ i18n/hreflang + GEO 就绪度 + P0/P1/P2 输出格式 |


English

Author: Iris (生姜iris) | Version: v1.0 (April 2026)

Core Philosophy

SEO and GEO are not two separate things — they're two sides of the same coin. Structured data serves both traditional search and AI search simultaneously.

Traditional SEO solves "being found." GEO solves "being cited by AI."

In 2026, search traffic splits into two entry points: Google/Bing rankings, and AI answers from ChatGPT/Perplexity/Claude. Your content must win on both battlefields.

SEO + GEO Dual-Engine Comparison

| Dimension | SEO (Traditional Search) | GEO (AI Search) |
|-----------|-------------------------|-----------------|
| Goal | Google/Bing rankings | Cited in AI answers |
| Core Signal | Backlinks + keyword match | Structured data + E-E-A-T authenticity |
| Content Format | H2/H3 hierarchy + long-tail coverage | Direct answers + comparison tables + FAQ Schema |
| Technical | XML Sitemap + Core Web Vitals | IndexNow + robots.txt allowing AI crawlers |
| Measurement | GA4 organic traffic + Ahrefs rankings | Manual check of AI citation frequency |

Five Core Principles

  1. Start from BOFU, work upward — High-intent keywords first (pricing, comparisons), then educational content. Highest-converting terms get priority.
  2. Authentic voice is the best E-E-A-T — Don't let AI write like AI. Use founder voice, real numbers, first-hand stories to build credibility.
  3. Structured = Citable — Key Stats tables, FAQ Schema, comparison matrices let AI engines directly extract and cite.
  4. IndexNow is GEO infrastructure — Push content updates to Bing/AI search in seconds, don't wait for crawlers.
  5. Comparison pages are SEO gold mines — One dedicated page per competitor, capturing decision-stage users.

Detailed Guides

| Topic | File | Description |
|:---|:---|:---|
| Keyword Funnel Strategy | references/en/keyword-strategy.md | TOFU/MOFU/BOFU keyword matrix & prioritization |
| SEO Technical Foundations | references/en/seo-foundations.md | Core page optimization, Schema, site architecture |
| GEO Optimization | references/en/geo-optimization.md | IndexNow, AI crawlers (search vs training split), Schema type quick-ref matrix, AI citation monitoring |
| Content Production SOP | references/en/content-sop.md | End-to-end workflow from topic selection to publishing |
| E-E-A-T Writing Voice | references/en/writing-voice.md | Founder voice system for authenticity signals |
| Comparison Page SOP | references/en/comparison-pages.md | Page structure, Schema, keyword embedding |
| Measurement Layer 🆕 | references/en/measurement-layer.md | Four core metrics, Article DB, GSC API, AI referral traffic, attribution traps |
| SEO Audit SOP 🆕 | references/en/seo-audit.md | Four-phase audit (Technical → On-Page → Content → Off-Page) + i18n/hreflang + GEO readiness + P0/P1/P2 output |


Key Results (Battle-Tested)

| Metric | Result |
|:---|:---|
| AFFiNE GitHub stars | 60,000+ (among fastest open-source projects to 10k pre-ChatGPT era) |
| GitHub Trending appearances | 28 times (Aug-Dec 2022) |
| Product Hunt #1 launches coached | 30+ |
| AI startups consulted | 150+ |
| Content cadence proven effective | 4 articles/week, KD 30-50 long-tail keywords |
| GSC impressions (one site, one month) | ~32,000 — full SOP at SEO-GEO-Agent-SOP.md |


🙏 Acknowledgments / 致谢

The Measurement Layer and SEO Audit references in this playbook draw on the open-source SEO/GEO skills published by kostja94 (Kostja Zhang) on ClawHub — specifically seo-monitoring, traffic-analysis, ai-traffic-tracking, seo-audit-strategies, internal-links, robots-txt, and schema-markup, all released under MIT-0. The four-core-metrics framework, the GA4 AI-source regex pattern, and the four-phase audit ordering are adapted from those skills, with battle-tested additions from Iris's own AFFiNE / Gingiris / 150+ consultation deployments.

Borrowing across the OSS skill ecosystem is exactly why MIT-0 exists. If you find this playbook useful, please also check out kostja94's full SEO/GEO skill suite on ClawHub — measurement, technical audit, and GA4 analytics depth that complement this playbook's strategy + voice + agent-operational layers.


Where to Find Us / 找到我们

  • 🎯 ClawHub (skills & playbooks): clawhub.ai/user/gingiris — 30+ skills covering launch, OSS, B2B SaaS, ASO, GTM, KOL outreach, user interview, SEO/GEO Agent SOP
  • 📚 HuggingFace (datasets): huggingface.co/Gingiris — 15 growth/marketing datasets including gingiris-skills, growth-tools, gingiris-seo-geo-agent
  • 🌐 Live blog & tool directory: gingiris.tools
  • ✍️ Dev.to syndicate: dev.to/iris1031