{
  "name": "麦一联盟搜索与AI爬虫身份核验方法",
  "description": "搜索与AI爬虫请求的名称观察、官方IP或DNS身份核验、最小化聚合与证据边界说明。",
  "version": "1.0",
  "updatedAt": "2026-07-28",
  "source": "https://www.mineaibd.com/research/crawler-verification/",
  "provenance": "按Google、Bing与OpenAI公开的爬虫验证说明整理。未公开可自动验证规则的平台仅记录User-Agent观察，不将其作为身份或收录证据。",
  "license": "可在注明“麦一联盟”、数据集名称、版本与下载网址的前提下引用；不得据此伪造背书或未发布的行业结论。",
  "recordCount": 5,
  "records": [
    {
      "id": "M01",
      "layer": "名称观察",
      "method": "按声明的User-Agent分类请求",
      "can_prove": "用于发现可能的搜索或AI访问",
      "cannot_prove": "不证明来源真实、收录、引用、推荐或转化"
    },
    {
      "id": "M02",
      "layer": "Google身份核验",
      "method": "反向DNS后再正向解析，或匹配Google公开爬虫IP范围",
      "can_prove": "来源通过官方规则时记录为已核验请求",
      "cannot_prove": "不能证明索引覆盖、排名或生成式回答引用"
    },
    {
      "id": "M03",
      "layer": "Bing身份核验",
      "method": "匹配Bing公开IP范围，或反向DNS后再正向解析",
      "can_prove": "来源通过官方规则时记录为已核验请求",
      "cannot_prove": "不能证明Bing收录、排名或流量"
    },
    {
      "id": "M04",
      "layer": "OpenAI身份核验",
      "method": "匹配OAI-SearchBot、GPTBot、ChatGPT-User公开IP范围",
      "can_prove": "区分搜索发现、训练用途和用户触发访问",
      "cannot_prove": "不能把任何一类访问写成ChatGPT回答推荐"
    },
    {
      "id": "M05",
      "layer": "最小化公开",
      "method": "仅发布日/周聚合计数、状态类别与方法版本",
      "can_prove": "保护访问来源与日志安全，便于长期复核",
      "cannot_prove": "不公开来源IP、完整User-Agent或原始服务器日志"
    }
  ]
}
