stagenth · 文档解析 is an MCP server with 4 tools an agent can call, among them doc_chunk, doc_images, doc_parse. It answers at https://stagenth.com/mcp/doc-parse/ over streamable HTTP, and it listed its tools to an anonymous client, so no sign-in was needed for that. Our probes reached it 6 of 6 times since 10 September 2026; the last check, on 19 October 2026, succeeded.
Parse PDF/Word/PPT/HTML to Markdown; tables as JSON, image extraction, RAG chunking, page ranges.
Description as the publisher wrote it.
| Tool | What it does |
|---|---|
doc_chunk | 把文档切成适合 RAG / 向量嵌入的语义块（按标题层级切，超长块按段落细分）。 返回 [{index, heading, text, chars}]，喂检索/嵌入无需再自己写切块逻辑。 Inputs: file_id, data_base64, filename, max_chars, page_range |
doc_images | 抽取文档内嵌的图片（PDF / .docx / .pptx），打包 ZIP 落文件中转站并返下载 URL。 过滤 <1KB 的图标噪声；多模态 AI 可配合 doc_parse 的文本一起理解图文文档。 Inputs: file_id, data_base64, filename, max_images |
doc_parse | 把 PDF / Word(.docx) / PPT(.pptx) / HTML / 纯文本 解析成干净 Markdown + 结构化（按实际解析页数计费，每页 1 credit）。 纯解析 + 格式转换、不生成任何 AI 内容；扫描件 OCR 暂未支持。 支持 page_range 只解析指定页码段（只按解析的页数扣费）。 Inputs: file_id, data_base64, filename, max_pages, page_range |
doc_tables | 抽取文档中的表格（PDF / .docx / .pptx），转成 headers+rows 的 JSON 可直接计算。 与 doc_parse 的区别：表格不再压平成 markdown 竖线文本，而是结构化行列， AI 可直接算合计/对比；PDF 用版面识别找表格。 Inputs: file_id, data_base64, filename, max_pages |
Every block below points at https://stagenth.com/mcp/doc-parse/. Pick your client and copy.
claude mcp add --transport http stagenth https://stagenth.com/mcp/doc-parse/
{
"mcpServers": {
"stagenth": {
"command": "npx",
"args": ["-y", "mcp-remote", "https://stagenth.com/mcp/doc-parse/"]
}
}
}{
"mcpServers": {
"stagenth": {
"url": "https://stagenth.com/mcp/doc-parse/"
}
}
}{
"servers": {
"stagenth": {
"type": "http",
"url": "https://stagenth.com/mcp/doc-parse/"
}
}
}{
"context_servers": {
"stagenth": {
"url": "https://stagenth.com/mcp/doc-parse/"
}
}
}{
"mcpServers": {
"stagenth": {
"serverUrl": "https://stagenth.com/mcp/doc-parse/"
}
}
}{
"mcpServers": {
"stagenth": {
"url": "https://stagenth.com/mcp/doc-parse/"
}
}
}name: stagenth · 文档解析
version: 0.0.1
schema: v1
mcpServers:
- name: stagenth · 文档解析
type: streamable-http
url: https://stagenth.com/mcp/doc-parse/{
"mcpServers": {
"stagenth": {
"httpUrl": "https://stagenth.com/mcp/doc-parse/"
}
}
}{
"mcpServers": {
"stagenth": {
"type": "http",
"url": "https://stagenth.com/mcp/doc-parse/"
}
}
}{
"mcpServers": {
"stagenth": {
"url": "https://stagenth.com/mcp/doc-parse/"
}
}
}goose://extension?url=https%3A%2F%2Fstagenth.com%2Fmcp%2Fdoc-parse%2F&type=streamable_http&id=stagenth&name=stagenth%20%C2%B7%20%E6%96%87%E6%A1%A3%E8%A7%A3%E6%9E%90
It answered our last check on 19 October 2026 in 1872 ms. Over 6 checks since 10 September 2026 it answered 6. Status here comes from Neuronto's own probes, not from the publisher.
Yes. It is published by stagenth.com and served from stagenth.com, the vendor's own domain: https://stagenth.com/mcp/doc-parse/.
4, as returned by the server's own tools/list on 11 September 2026: doc_chunk, doc_images, doc_parse, doc_tables.
It answered our last check on 19 October 2026 in 1872 ms. Over 6 checks since 10 September 2026 it answered 6.
Put the server under mcpServers in ~/.cursor/mcp.json with the URL https://stagenth.com/mcp/doc-parse/. The exact JSON is above, with the same setup for Claude Code, Claude Desktop, VS Code, Windsurf, Zed, Gemini CLI and the rest.
It listed its tools to an anonymous client, so reading the tool list needs no key. Individual tools can still ask for credentials in their own inputs.
Publisher: com.stagenth. Endpoint host: stagenth.com. ARD identifier: urn:air:com.stagenth:mcp:doc-parse.
First indexed 10 September 2026. The tools are exactly what the server returned; nothing on this page
is inferred from its name. Ranking in Neuronto search is never for sale.