AEO (Answer Engine Optimization) System
Get AI assistants — ChatGPT, Perplexity, Claude, Gemini — to recommend your brand when people ask purchase-intent questions.
What This Is
AEO is the discipline of optimizing for AI-powered answer engines the same way SEO optimizes for search engines. When someone asks Perplexity "what's the best magnesium supplement for sleep?" — AEO determines whether your brand gets named.
This skill gives an OpenClaw agent the ability to:
- 1. Audit a brand's current AEO infrastructure across all 7 layers
- Map which brands AI platforms recommend (and in what position) for any category
- Track position changes week over week
- Build the missing infrastructure (Answer Hub, brand-facts.json, schema, citations)
- Maintain the system with a weekly 90-minute protocol
When to Load This Skill
- - User asks to "run an AEO audit" for a brand or URL
- User asks "which brands are being recommended by AI for [category]?"
- User asks to "build an Answer Intent Map" for a category
- User asks to check a brand's Answer Hub, brand-facts.json, or schema markup
- User asks to track AI recommendation positions over time
- User asks to run the "weekly AEO maintenance protocol"
The 7-Layer AEO Framework
| Layer | Name | What It Is | Priority |
|---|
| 1 | Answer Intent Map | Spreadsheet of all purchase-intent queries + which brands AI recommends | Foundation |
| 2 |
Answer Hub | A long-form guide page that answers every key question in your category | High |
| 3 | Brand-Facts Page | Human-readable brand facts page (neutral, factual, cite-able) | High |
| 4 | brand-facts.json | Machine-readable brand data at
/.well-known/brand-facts.json | Medium |
| 5 | Schema Markup | Product, FAQ, and Organization structured data | Medium |
| 6 | Citation Network | Getting listed on the sources AI models actually cite | High |
| 7 | GPT Shopping | Google Merchant Center + review feed for AI shopping results | High |
Prerequisites
Required:
- -
PERPLEXITY_API_KEY — enables direct API queries (get free at perplexity.ai/settings/api) - Node.js v18+ (for the
answer-intent-map.js script)
Optional:
- -
OPENAI_API_KEY — enables ChatGPT query automation - INLINECODE4 — enables web searches for infrastructure checks
Without API keys: The skill runs in "manual-assist" mode — generates the queries, provides a blank log template, and analyzes results you paste in.
Core Workflows
Workflow 1: Full AEO Audit
Trigger: "Run an AEO audit for [brand URL]"
Steps:
- 1. Fetch and analyze the brand's website for AEO infrastructure:
- Check for Answer Hub page (
/guides/ or similar long-form page)
- Check for Brand-Facts page (
/brand-facts)
- Check for machine-readable data (
/.well-known/brand-facts.json)
- Audit schema markup on product pages (via Rich Results API or web_fetch)
- Check for a Wikidata entry
- Check Google Merchant Center eligibility signals
- 2. Score each of the 7 layers (0–3 scale):
-
0 = Doesn't exist
-
1 = Exists but incomplete or outdated
-
2 = Exists, functional, minor gaps
-
3 = Complete, current, optimized
- 3. Generate a gap analysis report with:
- Current score per layer
- Priority order for implementation
- Specific action items for each missing layer
Output: Markdown report saved as aeo-audit-[brand]-[date].md
Workflow 2: Answer Intent Map
Trigger: "Build an Answer Intent Map for [category]" or run INLINECODE9
Steps:
- 1. Generate query list from four types:
-
Category queries: "best [product] for [use case]" (10–15 queries)
-
Comparison queries: "[brand] vs [competitor]" (10 queries)
-
Brand queries: "is [brand] worth it" (5 queries)
-
Educational queries: "does [ingredient] help with [condition]" (10 queries)
- 2. For each query, query available platforms:
- Perplexity API (structured JSON response with citations)
- OpenAI API (text response — brand names extracted by parser)
- Browser fallback for Claude and Gemini
- 3. Parse responses to extract:
- Brand names mentioned (position 1, 2, 3)
- Source URLs cited
- Key verbatim quotes
- 4. Write results to JSON data file + Markdown summary report
Output: answer-intent-map-[category]-[date].json + .md summary
Run the script:
node scripts/answer-intent-map.js \
--category "magnesium supplements" \
--brand "MyBrand" \
--queries 20
# Or with a config file:
node scripts/answer-intent-map.js --config ./aeo-config.json
Workflow 3: Weekly Maintenance Protocol
Trigger: "Run weekly AEO maintenance" or scheduled cron
Steps:
- 1. Load the brand's Answer Intent Map (top 15 priority queries)
- Query ChatGPT and Perplexity for each priority query in fresh sessions
- Compare results against previous week's log (detect position changes)
- Generate maintenance report:
- Position changes (up/down/new competitors)
- New sources being cited this week
- Recommended Answer Hub updates
- 5. Check
brand-facts.json for stale lastUpdated timestamp - Check Google Merchant Center for disapprovals (via browser if needed)
Output: INLINECODE14
Use the checklist: templates/weekly-maintenance-checklist.md
Workflow 4: Citation Network Analysis
Trigger: "Analyze AEO citations for [category]"
Steps:
- 1. Run 20 category queries via Perplexity API (citations returned directly)
- Extract all unique source URLs from responses
- Group and count by domain
- Identify top 10 most-cited external sources in the category
- Generate outreach priority list
Output: Citation analysis report with target sites ranked by citation frequency
Workflow 5: Infrastructure Build
Trigger: "Build AEO infrastructure for [brand]" or "Set up brand-facts.json"
Steps:
- 1. Ask for brand details (or load from
aeo-config.json) - Generate from templates:
-
brand-facts.json →
templates/brand-facts.json (fill placeholders)
- Answer Hub page →
templates/answer-hub-template.md
- Schema markup snippet (JSON-LD for product pages)
- 3. Provide implementation instructions per asset
Configuration
Create aeo-config.json in your working directory:
CODEBLOCK1
Output Files
| File | Description |
|---|
| INLINECODE21 | Infrastructure audit report |
| INLINECODE22 |
Raw AI query results |
|
answer-intent-map-[category]-[date].md | Human-readable competitive summary |
|
aeo-weekly-report-[date].md | Weekly maintenance report |
|
citation-analysis-[category]-[date].md | Citation network analysis |
Usage Examples
CODEBLOCK2
Platform Limitations
| Platform | Access | Notes |
|---|
| Perplexity | API (structured, reliable) | Returns citations directly |
| ChatGPT |
API via OpenAI | Text parsing required for brand extraction |
| Claude | Browser required | Generate queries + blank log; agent uses browser |
| Gemini | Browser required | Generate queries + blank log; agent uses browser |
For Claude/Gemini: the skill generates the query list and a blank log template; use the browser tool to collect results.
Rate limits: Perplexity free tier ≈ 20 requests/minute. For 50+ queries, add --delay 3000 to the script.
File Structure
CODEBLOCK3
AEO System v1.0 — February 2026
A product by Carson Jarvis (@CarsonJarvisAI)AEO（答案引擎优化）系统
当人们提出购买意向问题时，让AI助手——ChatGPT、Perplexity、Claude、Gemini——推荐您的品牌。
这是什么
AEO是一门针对AI驱动的答案引擎进行优化的学科，就像SEO针对搜索引擎进行优化一样。当有人问Perplexity哪种镁补充剂最适合睡眠？时——AEO决定了您的品牌是否会被提及。
这项技能赋予OpenClaw代理以下能力：
- 1. 审计品牌在所有7个层面的当前AEO基础设施
- 映射AI平台推荐哪些品牌（以及排名位置）用于任何品类
- 追踪每周排名变化
- 构建缺失的基础设施（答案中心、brand-facts.json、结构化数据、引用源）
- 维护系统，每周执行90分钟协议
何时加载此技能
- - 用户要求对品牌或URL运行AEO审计
- 用户询问AI为[品类]推荐了哪些品牌？
- 用户要求为某个品类构建答案意图地图
- 用户要求检查品牌的答案中心、brand-facts.json或结构化数据标记
- 用户要求追踪AI推荐排名随时间的变化
- 用户要求运行每周AEO维护协议
7层AEO框架
| 层级 | 名称 | 说明 | 优先级 |
|---|
| 1 | 答案意图地图 | 所有购买意向查询的电子表格 + AI推荐的品牌 | 基础 |
| 2 |
答案中心 | 回答品类中每个关键问题的长篇指南页面 | 高 |
| 3 | 品牌事实页面 | 人类可读的品牌事实页面（中立、事实性、可引用） | 高 |
| 4 | brand-facts.json | 机器可读的品牌数据，位于/.well-known/brand-facts.json | 中 |
| 5 | 结构化数据标记 | 产品、FAQ和组织结构化数据 | 中 |
| 6 | 引用网络 | 出现在AI模型实际引用的来源列表中 | 高 |
| 7 | GPT购物 | 面向AI购物结果的Google Merchant Center + 评论数据源 | 高 |
前提条件
必需：
- - PERPLEXITYAPIKEY — 启用直接API查询（在perplexity.ai/settings/api免费获取）
- Node.js v18+（用于answer-intent-map.js脚本）
可选：
- - OPENAIAPIKEY — 启用ChatGPT查询自动化
- BRAVEAPIKEY — 启用基础设施检查的网络搜索
没有API密钥： 技能以手动辅助模式运行——生成查询、提供空白日志模板，并分析您粘贴的结果。
核心工作流程
工作流程1：完整AEO审计
触发： 对[品牌URL]运行AEO审计
步骤：
- 1. 获取并分析品牌的网站以检查AEO基础设施：
- 检查答案中心页面（/guides/或类似的长篇页面）
- 检查品牌事实页面（/brand-facts）
- 检查机器可读数据（/.well-known/brand-facts.json）
- 审计产品页面上的结构化数据标记（通过Rich Results API或web_fetch）
- 检查Wikidata条目
- 检查Google Merchant Center资格信号
- 2. 对7个层级分别评分（0–3分制）：
-
0 = 不存在
-
1 = 存在但不完整或过时
-
2 = 存在、功能正常、有少量差距
-
3 = 完整、最新、已优化
- 3. 生成差距分析报告，包含：
- 每个层级的当前得分
- 实施优先级顺序
- 每个缺失层级的具体行动项
输出： Markdown报告，保存为aeo-audit-[品牌]-[日期].md
工作流程2：答案意图地图
触发： 为[品类]构建答案意图地图或运行scripts/answer-intent-map.js
步骤：
- 1. 从四种类型生成查询列表：
-
品类查询： 最佳[产品]用于[使用场景]（10–15个查询）
-
对比查询： [品牌] vs [竞争对手]（10个查询）
-
品牌查询： [品牌]值得买吗（5个查询）
-
教育查询： [成分]对[状况]有帮助吗（10个查询）
- 2. 对每个查询，查询可用平台：
- Perplexity API（带引用的结构化JSON响应）
- OpenAI API（文本响应——由解析器提取品牌名称）
- Claude和Gemini的浏览器回退方案
- 3. 解析响应以提取：
- 提及的品牌名称（位置1、2、3）
- 引用的来源URL
- 关键逐字引用
- 4. 将结果写入JSON数据文件 + Markdown摘要报告
输出： answer-intent-map-[品类]-[日期].json + .md摘要
运行脚本：
bash
node scripts/answer-intent-map.js \
--category 镁补充剂 \
--brand 我的品牌 \
--queries 20
或使用配置文件：
node scripts/answer-intent-map.js --config ./aeo-config.json
工作流程3：每周维护协议
触发： 运行每周AEO维护或定时任务
步骤：
- 1. 加载品牌的答案意图地图（前15个优先查询）
- 在新的会话中查询ChatGPT和Perplexity的每个优先查询
- 将结果与上周日志进行比较（检测排名变化）
- 生成维护报告：
- 排名变化（上升/下降/新竞争对手）
- 本周被引用的新来源
- 推荐的答案中心更新
- 5. 检查brand-facts.json中lastUpdated时间戳是否过时
- 检查Google Merchant Center是否有拒绝项（必要时通过浏览器）
输出： aeo-weekly-report-[日期].md
使用清单： templates/weekly-maintenance-checklist.md
工作流程4：引用网络分析
触发： 分析[品类]的AEO引用
步骤：
- 1. 通过Perplexity API运行20个品类查询（直接返回引用）
- 从响应中提取所有唯一的来源URL
- 按域名分组并计数
- 识别品类中被引用最多的前10个外部来源
- 生成外展优先级列表
输出： 引用分析报告，按引用频率排序的目标网站
工作流程5：基础设施构建
触发： 为[品牌]构建AEO基础设施或设置brand-facts.json
步骤：
- 1. 询问品牌详细信息（或从aeo-config.json加载）
- 从模板生成：
- brand-facts.json → templates/brand-facts.json（填充占位符）
- 答案中心页面 → templates/answer-hub-template.md
- 结构化数据标记片段（产品页面的JSON-LD）
- 3. 为每个资产提供实施说明
配置
在工作目录中创建aeo-config.json：
json
{
brandName: 您的品牌名称,
brandUrl: https://yourbrand.com,
category: 镁补充剂,
priorityQueries: [
最佳睡眠镁补充剂,
最佳甘氨酸镁补充剂,
缓解焦虑的镁补充剂
],
competitors: [
竞争对手品牌A,
竞争对手品牌B,
竞争对手品牌C
],
answerHubUrl: https://yourbrand.com/guides/best-magnesium-supplements-2026,
brandFactsJsonUrl: https://yourbrand.com/.well-known/brand-facts.json
}
输出文件
| 文件 | 描述 |
|---|
| aeo-audit-[品牌]-[日期].md | 基础设施审计报告 |
| answer-intent-map-[品类]-[日期].json |
原始AI查询结果 |
| answer-intent-map-[品类]-[日期].md | 人类可读的竞争摘要 |
| aeo-weekly-report-[日期].md | 每周维护报告 |
| citation-analysis-[品类]-[日期].md | 引用网络分析 |
使用示例
完整基础设施审计
对mybrand.com运行AEO审计
构建竞争情报
为镁补充剂品类构建答案意图地图
快速排名检查
检查mybrand.com是否被Perplexity推荐为最佳睡眠镁补充剂
每周维护
为我的品牌运行每周AEO维护协议
引用分析
Perplexity在推荐胶原蛋白补充剂时最常引用哪些来源？
生成brand-facts.json
