SuperAPI AI Directory
Partner with us中文AI newsFavorites
Skillsgeo-optimizer

geo-optimizer

Agents & models

View repository

Updated 10-15 23:59

About this skill

GEO Optimizer

A full-process AI search optimization tool, based on a 10-step GEO hands-on framework, that helps brand content get cited by AI engines.

Workflow Decision Tree

Choose an execution path based on user needs:

用户提供 URL 要求优化？
  → 模块 E：内容差距分析（调用 /geo-content-optimizer）

用户要求生成 Schema/JSON-LD？
  → 模块 A：结构化数据生成

用户提供文本要求检测质量？
  → 模块 B：内容 AI 可读性检测

Does the user want to plan content architecture? → Module C: Pillar Page Cluster Planning

Does the user want to test/monitor AI visibility? → Module D: Visibility Testing and Reporting

Is this the user's first time / unsure where to start? → Execute the "First Full GEO Setup" process (see below)


---

## 首次完整 GEO 设置流程

适用于从未做过 GEO 优化的用户，按顺序执行：

1. 收集品牌信息（名称、行业、核心业务、目标市场）
2. 执行模块 A1：生成 Organization Schema
3. 引导用户创建 ≥10 个 FAQ 问答对
4. 执行模块 A2：生成 FAQ Schema
5. 执行模块 A3：为已有文章生成 Article Schema
6. 执行模块 B：检测关键页面的 AI 可读性
7. 执行模块 C：规划 Pillar Page 内容集群
8. 执行模块 D1：首次 AI 可见度基线测试
9. 汇总输出完整 GEO 部署清单

---

## 模块 A：结构化数据生成

使用 `scripts/schema_generator.py` 或参考 `references/schema-templates.md` 生成 JSON-LD。

### A1. Organization 实体标记

收集以下信息后生成：
- 品牌名称、一句话描述、成立年份
- 目标市场 (areaServed)
- 核心领域 (knowsAbout, 3-5 个)
- 提供的产品/服务描述

```bash
python scripts/schema_generator.py --type organization --name "品牌名" --description "描述" --knows-about "领域1" "领域2" --offers "产品描述"

A2. FAQ Schema

Require the user to provide ≥10 Q&A pairs. Each Q&A should:

  • The question uses natural spoken language (how users would ask AI)
  • The answer contains specific data, cases, and actionable suggestions
  • The answer length is 80-200 characters
python scripts/schema_generator.py --type faq --input faq_pairs.json

Input file format:

[
  {"question": "新手第一步应该做什么？", "answer": "具体答案..."},
  {"question": "A 和 B 有什么区别？", "answer": "具体答案..."}
]

A3. Article Schema

Generate TechArticle or Article Schema for each article.

A4. E-E-A-T Author Markup

Generate a combined Person + Article @graph markup. Requirements:

  • Author byline + bio (cannot be "Anonymous"/"Site Editor")
  • List social accounts (sameAs)
  • List credentials/accomplishments (hasCredential)

A5. Citation Markup

Add citation Schema for articles that cite authoritative sources.

For Schema templates, see：references/schema-templates.md

Module B: Content AI Readability Detection

Use scripts/readability_checker.py to analyze text.

python scripts/readability_checker.py --text "要检测的文本内容"
python scripts/readability_checker.py --file article.txt

Detection dimensions:

  1. Marketing word density — "innovative/leading/excellent/comprehensive/one-stop/ultimate/disruptive/best", etc.; the fewer the better
  2. Average sentence length — ideally 15-25 characters; >35 characters is hard for AI to understand
  3. Direct questions — sentences containing "？/?"; AI prefers content with questions
  4. Data citations — specific numbers/percentages/amounts; AI trusts content containing data more
  5. Overall score — High/Medium/Low

Optimization suggestion rules:

  • Marketing words > 3 → suggest replacing with specific factual descriptions
  • Sentence length > 25 characters → suggest splitting long sentences
  • No questions → suggest adding 1-2 direct questions
  • No data → suggest adding industry data or proprietary statistics

Module C: Pillar Page Cluster Planning

Design content cluster structure

1 core topic page + 5-10 subtopics, linked together through internal links.

Collect user input:

  • Core topic (Pillar Page title)
  • Subtopic list (5-10 related topics)
  • The relationship between each subtopic and the core topic

Generate content:

  • Cluster HTML code block (placed at the end of each sub-article)
  • Sitemap JSON structure
  • Internal linking strategy suggestions

For templates, see：assets/pillar-template.html

Module D: Visibility Testing and Reporting

D1. AI Visibility Testing (Monthly)

Use scripts/visibility_tester.py to call the Perplexity API.

python scripts/visibility_tester.py --brand "品牌名" --keywords "关键词1" "关键词2" --api-key "pplx-xxx"

Require the user to provide a Perplexity API Key (get it at:https://docs.perplexity.ai/）。

The script will:

  1. Construct queries with multiple keywords
  2. Detect whether the brand name appears in the AI response
  3. Extract citation context
  4. Results are automatically appended to the history JSON file

When there is no API Key: Provide manual testing guidance (search the target keywords in ChatGPT/Perplexity/Gemini and observe whether the brand is mentioned).

D2. Quarterly GEO Report

Use scripts/geo_report_generator.py。

python scripts/geo_report_generator.py --history visibility_history.json

The report includes:

  • Citation rate trend (first half vs second half)
  • Trend direction (rising/falling/flat)
  • Action recommendations (automatically generated based on the trend)

Module E: Content Gap Analysis

When the user provides a URL and requests optimization, call the existing /geo-content-optimizer skill。

Trigger words: "optimize this page", "analyze this URL", "GEO analysis", the user provides a URL and requests improvement.

Execution method:/geo-content-optimizer <url>

This skill will automatically complete: fetch title → Google query expansion → AI Overview retrieval → comparative analysis → optimization recommendation report.

30-Minute Quick Implementation Checklist

Step Time Action
Entity markup 5 min Module A1: Generate Organization JSON-LD
FAQ Schema 10 min Module A2: Write 10 FAQs + generate markup
Conversational check 5 min Module B: Run readability check
Citation markup 5 min Module A5: Add citation markup
Full Schema deployment 10 min Module A3/A4: Article + Brand + Author
Pillar cluster 30 min Module C: Design 1+5 content structure
Visibility test 5 min/month Module D1: Run API script
Quarterly report 10 min/quarter Module D2: Run comparative analysis

Reference Materials