GhostAPI

Turn any public website into structured JSON using AI and natural language queries. **Tools:** - `ghostapi_fetch` — scrape any public URL, extract structured data, get ethical semantic evaluation - `ghostapi_read` — read extraction history for a URL, zero LLM tokens Built for AI agents, Claude workflows, and LLM applications that need real-world web data without maintaining scrapers. **Get your API key:** rapidapi.com (search GhostAPI)

该来源不提供完整文件导出（国内平台多为平台内托管），仅存元数据与原链

接入信息

传输形态: http
鉴权方式: 需 API Key（需要配置：Api）
端点: https://ghostapi--dvd-pzz05.run.tools

该服务需要凭证，请按官方文档申请后替换占位符

{
  "mcpServers": {
    "GhostAPI": {
      "headers": {
        "Authorization": "Bearer \u003cYOUR_KEY\u003e"
      },
      "url": "https://ghostapi--dvd-pzz05.run.tools"
    }
  }
}

能力清单

工具

说明

ghostapi_fetch

Fetch any public URL and extract structured JSON using a natural language query. Results are cached automatically — repeated requests return instantly at no extra cost. Every live extraction includes a semantic context evaluation scoring the content on cooperation index (0-1), social utility (0-1), and environmental impact. Supports news sites, e-commerce, company pages, job listings, public directories. Does not support login-required pages or Cloudflare-protected sites.

ghostapi_read

Read the full extraction history for a URL from the HistoricalData database. Zero LLM tokens consumed — pure database read. Returns timestamped records with extracted JSON and semantic context scores. Use to analyze how a source changes over time, track cooperation index trends, compare social utility across multiple captures, or build time-series insights from accumulated web data.

纠错与举报（发现条目失效、署名有误或涉及侵权？）

提交举报 / 纠错

侵权举报经核验成立后，我们会即时下线该条目并删除已存的内容副本。