网页抓取 MCP 推荐

Token-first headless browser for AIs. Extracts a tiny semantic graph, cutting HTML bloat by 95%.

截图/PDF/HTML 渲染 API；需 API 密钥或 x402，已关闭免密钥访问。

Parsed SEC events by CIK/CUSIP: 8-K items, Form 144, Form D, FTD. Not EDGAR HTML.

597 motion/3D UI patterns as ready HTML/CSS/JS, served over MCP. The entire library is free.

由 HTML 生成带旁白的演示文稿，可轮询构建状态、列出租户清单，并把其中一份读回为文本。

Google 评论 API——从 Google、Trustpilot、TripAdvisor、G2 等 50 多个平台抓取评论。

AI 设计工作室：从简报出发产出方向板、HTML 原型、图片与视频，使用 OAuth 登录。

通过 MCP 从 FRED API 检索约 80 万条美联储经济时间序列。

Deploy HTML or dist/ to a live HTTPS URL. No account. Drop, MCP, or CLI.

26 pay-per-call tools for agents: scrape, verify, guards, crypto. x402 USDC on Base.

按工种或地点检索美国持牌承包商，获取档案与评价，并提交线索。

Search people open to work and fetch their machine-readable cv.json, from any AI assistant.

通过一个远程 MCP 工具，把公开 HTML 或公开 URL 转成 token 密度更高的 Markdown。

Search song lyrics on lyrics.com by word or by title, and fetch full lyrics. No API key required.

Convert web pages or HTML to clean Markdown for LLM context. Local, no external API calls.

Paid x402 MCP utilities for AI agents on Base: JSON, HTML, CSV, hashing, attestations.

Shared copies of public web pages for AI agents. Search stored pages or fetch a URL.

文档问答：检索 169 篇数据与 AI 指南，任意页面可取 Markdown 格式内容；只读且无需密钥。

AI visibility checker: GPTBot vs OAI-SearchBot, robots.txt audit, raw-HTML schema. 165 crawlers.

x402 pay-per-call APIs for AI agents on Base: site audits, llms.txt, wallet watch, prices, fetch.

检索经认证的 Claude Code 插件与技能，获取可移植的 SKILL.md 源码；只读。

基于我们自建抓取索引的实时网页搜索、图片搜索、主题过滤与全文抓取。

通过 FileChanger API 在文本类文档格式（Markdown、HTML 等）之间转换。

搜索并获取 AgendaForge 的公开文档与营销内容，无需鉴权。

Market evidence with receipts: every claim resolves to a real stored record you can fetch back.

跨客户端 ECharts 图表 MCP server：数据 -> 可交互 HTML 与内联图片

Render-and-verify API: HTML/CSS to image or PDF, screenshot any URL, confirm the text rendered.

HSM 支撑的 AI 智能体密钥库（JIT 即时取密），附带提示注入与威胁扫描。

MCP server for Singapore Statutes Online: browse, fetch, and cite Singapore Acts.

MCP server for Laws of Malaysia Online: fetch and cite Malaysian principal Acts.

MCP server for Japan's e-Gov law search API: search, fetch, and cite national legislation.

在 21 个邮件客户端中分析、审计、修复与对比 HTML、MJML、Maizzle 和 React Email。

Web toolkit for AI agents: read any URL as Markdown, screenshot, extract data, or crawl a site.

发布众包内容采集任务，并经 FirstHandAPI 拉取审核通过的文件。

The documentation, as a tool your agent can call: 950+ AI-dev guides. Search + fetch tools.

从任意 MCP 兼容 AI 智能体将 HTML 发布到私有、不可猜测、可原地替换的 URL。

X/Twitter 自动化 MCP：抓取、发帖、排期、分析与互动，无需 API 密钥。

浏览社区 3D 创作画廊，按查看 URL 获取作品，并提交作品。

WhatsApp 资料数据 API，可编程获取用户的头像等资料。

Instagram 公开数据抓取 API，可搜索用户、帖子、话题标签、地点等。

Cloud PDF generation from HTML, CSS and XSL-FO, with PDF/A and PDF/UA support.

用模板+JSON 生成 PDF/DOCX/XLSX/PPTX，将 Office/HTML/MD 转为 PDF，支持通用模板。

Web toolkit for AI agents: read any URL as Markdown, screenshot, extract data, or crawl a site.

Publish an HTML presentation — a slide deck or scrolling document — at a shareable link.

Turn an HTML template plus JSON into a finished PDF: a quote, invoice, work order or certificate.

Deploy HTML from any agent: POST markup, get a live URL. Static hosting API with MCP tools.

Browserbase MCP server: cloud browser sessions, agents, contexts, fetch, search, functions

Stdio + HTTP MCP gateway: SSRF-hardened web_fetch, web_search, metis_verify, market_search.