@hauntapi/mcp-server
Haunt API MCP 服务器
从任何 URL 提取整洁、结构化的数据 —— 直接在 Claude、Cursor 或任何兼容 MCP 的 AI 中使用。
快速开始
添加到您的 Claude Desktop 配置中(Mac 上位于 ~/Library/Application Support/Claude/claude_desktop_config.json):
{
"mcpServers": {
"haunt": {
"command": "npx",
"args": ["-y", "@hauntapi/mcp-server"],
"env": {
"HAUNT_API_KEY": "your-api-key"
}
}
}
}在 hauntapi.com 获取免费 API 密钥 —— 每月 100 次请求,无需信用卡。
Related MCP server: Crawl4AI MCP Server
工具
extract_url
从任何 URL 提取整洁的数据。自动处理 JavaScript 渲染和 Cloudflare 绕过。
Extract the article content from https://example.com/blog/postextract_batch
同时从多个 URL 提取数据(最多 10 个)。
Extract the title and price from these product pages: [url1, url2, url3]extract_metadata
获取元数据:标题、描述、Open Graph 标签、Twitter 卡片、规范 URL。
Get the Open Graph metadata for https://example.com定价
免费版:每月 100 次请求
专业版:$0.01/次请求(按量付费)
链接
API 文档:https://hauntapi.com/docs
获取 API 密钥:https://hauntapi.com
Available Tools
5 toolsextract_articleA
Extract the main article content from a news article, blog post, or editorial page. Returns a JSON object with: title (string), body (string, full article text), author (string or null), and published_date (string or null). Use this instead of extract_url when you specifically need article content, it is a focused shortcut with consistent article fields. Read-only, makes no changes to any external system. Requires HAUNT_API_KEY environment variable. Free tier: 1,000 credits/month. Returns an error if rate limit, credit quota, or API key is invalid.
Inputs: url (required).
extract_markdownA
Return clean Markdown from a permitted public web page for agents, RAG ingestion, notes, or .md files. This is a low-cost non-LLM output mode when the page can be fetched cleanly. Blocked, login-required, CAPTCHA-gated, paywalled, and too-thin pages return explicit errors instead of fabricated Markdown. Read-only, makes no changes to any external system. Requires HAUNT_API_KEY environment variable.
Inputs: url (required).
extract_metadataA
Extract page metadata from a public or authorised URL: title, meta description, Open Graph tags (og:title, og:description, og:image, og:url), Twitter Card tags, canonical URL, and any other meta information present. Returns a JSON object with all discovered meta tags grouped by type. Use this instead of extract_url when you only need metadata, it is faster and returns a consistent schema. Read-only, makes no changes to any external system. Requires HAUNT_API_KEY environment variable. Free tier: 1,000 credits/month. Returns an error if rate limit, credit quota, or API key is invalid.
Inputs: url (required).
extract_urlA
Extract structured data from permitted public web pages by providing a URL and describing what you want. Returns clean JSON with exactly the fields you asked for by default. Can also return clean Markdown or raw HTML when response_format is set. Uses supported fetch paths for JavaScript-heavy pages and returns explicit error signals when blocked. It does not solve CAPTCHA, access login/paywall-only pages, or circumvent anti-bot controls. This is the general-purpose extraction tool. Use extract_markdown for LLM/RAG-ready Markdown, extract_article for full article content, or extract_metadata for page meta tags instead, they are optimised shortcuts. Read-only, makes no changes to any external system. Requires HAUNT_API_KEY environment variable. Free tier: 1,000 credits/month. Returns an error if rate limit, credit quota, or API key is invalid.
Inputs: url (required), prompt (required), response_format.
try_demo_extractA
Try Haunt's fixed demo extraction without an API key. Use this first when the user wants to verify the MCP package or see the activation path before signup. Returns demo, docs, signup, pricing, and free-tier links. No API key required. Free tier: 1,000 credits/month.
Inputs: no named parameters.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 5 tools
Each tool targets a distinct extraction purpose: article content, markdown, metadata, general extraction, and a demo without API key. Descriptions explicitly guide when to use which, eliminating ambiguity.
All tool names follow a consistent verb_noun snake_case pattern (extract_*, try_demo_extract). No mixed conventions or vague naming.
5 tools is well-scoped for a web extraction server, covering the main use cases without unnecessary redundancy.
Covers the core extraction needs with specialized shortcuts and a general-purpose tool. Minor gap: no tool for extracting specific structured elements like tables, but extract_url can handle it.
Maintenance
Related MCP Connectors
Web scraping MCP server: any URL to Markdown or JSON, with JavaScript rendering and batch jobs.
Web scraping MCP server — scrape, extract structured data, screenshot any site with anti-bot bypass.
Web scraping, browser automation, screenshots and AI data extraction for AI agents via MCP.
Web scraping for AI agents. Converts URLs to clean, LLM-ready Markdown with anti-bot bypass.
Related MCP Servers
- AlicenseAqualityDmaintenanceAn MCP server that fetches web pages and extracts clean, AI-friendly Markdown content using Mozilla Readability. It provides secure web access for LLMs with built-in SSRF protection and automated content cleaning for improved context retrieval and summarization.180 npmMIT
- FlicenseNot gradedqualityDmaintenanceA locally-hosted MCP server that provides AI assistants with advanced web crawling capabilities, including structured data extraction, deep site crawling, and page screenshots. It enables users to convert single or multiple URLs into clean Markdown content for processing by LLMs without requiring external API keys for basic features.-
- AlicenseNot gradedqualityDmaintenanceAn MCP server for intelligent web content extraction from JavaScript-heavy sites using single-file and trafilatura. It enables AI agents to fetch, render, and paginate through clean article content and metadata.23MIT
- AlicenseAqualityBmaintenanceAn MCP server that enables AI assistants to fetch web content in multiple formats (HTML, JSON, text, Markdown) with intelligent content extraction, chunk management, and browser automation support.6165 npm15MIT