Pagemd
Read any public web page as clean Markdown for LLMs, with its metadata. Paid per call, x402.
我该使用它吗
质量与安全性
基于对工具定义和协议合规性的自动分析。
上下文开销
这是每次将服务器的工具加载到模型上下文窗口时所消耗的大致 token 数。数值越高，可用于其他任务的注意力就越少。
安装
一键安装
将以下内容添加到你的 `claude_desktop_config.json` 文件中：
{
"mcpServers": {
"pagemd": {
"url": "https://pagemd.kaneky.dev/mcp"
}
}
}远程端点
https://pagemd.kaneky.dev/mcpstreamable-http它能做什么
工具清单
工具（1）
🟢read_page(url)
Read a public web page as clean Markdown for an LLM: the main content only (navigation, footers, sidebars, cookie banners, forms and comments removed), with headings, lists, GFM tables, code blocks, and links and images as absolute URLs. Also returns the title, canonical and final URL, language, byline and published date when the page states them, the meta/OpenGraph description, every JSON-LD block, and a word count. Markdown is capped at 60,000 characters (truncated flag). JavaScript is never run: a client-rendered page with no server HTML content is refused and not charged.
输入模式
{
"type": "object",
"properties": {
"url": {
"description": "The page: any public http or https address serving HTML.",
"format": "uri",
"maxLength": 2048,
"type": "string"
}
},
"required": [
"url"
]
}社区
证据