mcpField
超
超能文献MCP(AI文档翻译+文献检索)

超能文献MCP(AI文档翻译+文献检索)

Local
github.com/WildDataXSearch257Updated 29w ago

About 超能文献MCP(AI文档翻译+文献检索)

Suppr is an MCP server for researchers and AI assistants that offers document translation and semantic PubMed literature search. It translates common document formats while preserving formatting, and returns bibliographic details such as DOI and abstracts.

Key features

  1. 01Multi-format document translation
  2. 02Translation across 13 languages
  3. 03Preserves document formatting
  4. 04Semantic PubMed search
  5. 05Translation task status checks

Use cases

  1. 01Translate academic PDFs or Office documents
  2. 02Find biomedical papers with natural-language queries
  3. 03Check translation task progress

MCP support

As an MCP server, it provides compatible MCP clients with document translation management and PubMed literature search tools.

Configuration

Local server config is the same for every client. Copy it into your client’s MCP config file.

mcp.json · Universal config
{
"mcpServers": {
"suppr": {
"command": "npx",
"args": [
"-y",
"suppr-mcp"
],
"env": {
"SUPPR_API_KEY": "<SUPPR_API_KEY>"
}
}
}
}
Run commandnpx -y suppr-mcp
Environment variables to set
SUPPR_API_KEYSecret

Local servers need a runtime on your machine (e.g. Node.js 18+). Config file locations differ by client; see your client’s docs.

Packages
npmsuppr-mcp

Tools

4

Saved tool list · updated 2w ago

ParameterTypeRequiredDescription
file_pathstringNoLocal file path to translate (mutually exclusive with file_url)
file_urlstringNoDocument URL to translate (mutually exclusive with file_path)
from_langstringNoSource language code (optional, auto-detect if not specified)
optimize_math_formulabooleanNoOptimize math formulas (PDF only)
to_langstringYesTarget language code (required), e.g., en, zh, ko, ja

Details

From the project README

Suppr MCP - Usage Guide | 提供文档翻译及中文搜Pubmed的MCP服务 | Suppr超能文献

Suppr MCP Server

Suppr (超能文献) is an AI-powered academic tools platform by WildData. This MCP server brings document translation and literature search capabilities to AI assistants.

🌐 AI Document Translation — Translate PDF, Word (.docx), Excel (.xlsx), PowerPoint (.pptx), TXT, and HTML documents across 13 languages. Preserves original formatting. Automatic source language detection.

🔬 PubMed Academic Search — Semantic literature discovery across millions of biomedical research papers. Returns structured metadata: DOI, PMID, journal impact factor, citation counts, author affiliations, abstracts, and direct paper links.

🤖 MCP Compatible — Works with Claude Desktop, Cursor, Windsurf, and any Model Context Protocol client.

npm version License: MIT

Install

```bash npx suppr-mcp ```

Quick Start

1. 安装

全局安装:

npm install -g suppr-mcp

或者使用 npx(无需安装):

npx suppr-mcp
2. 获取 API Key

访问 Suppr API 获取您的 API 密钥。

3. 配置环境变量
export SUPPR_API_KEY=your_api_key_here
4. 在 MCP 客户端中使用
Claude Desktop 配置

编辑 ~/Library/Application Support/Claude/claude_desktop_config.json(macOS)或相应配置文件:

{
  "mcpServers": {
    "suppr": {
      "command": "npx",
      "args": ["-y", "suppr-mcp"],
      "env": {
        "SUPPR_API_KEY": "your_api_key_here"
      }
    }
  }
}

或者使用全局安装:

{
  "mcpServers": {
    "suppr": {
      "command": "suppr-mcp",
      "env": {
        "SUPPR_API_KEY": "your_api_key_here"
      }
    }
  }
}

可用工具

1. create_translation - 创建翻译任务

创建文档翻译任务。

参数:

  • file_path (file_path 和 file_url 二选一): 源文件路径
  • file_url (file_path 和 file_url 二选一): 要翻译的文档 URL
  • to_lang (必填): 目标语言代码
  • from_lang (可选): 源语言代码(默认自动检测)
  • optimize_math_formula (可选): 优化数学公式(仅 PDF)

示例:

{
  "file_url": "https://example.com/document.pdf",
  "to_lang": "en",
  "from_lang": "zh",
  "optimize_math_formula": true
}

返回:

{
  "task_id": "02a6c6d1-3f70-4a5a-80bc-971d53a37bb1",
  "status": "INIT",
  "consumed_point": 453,
  "source_lang": "zh",
  "target_lang": "en",
  "optimize_math_formula": true
}
2. get_translation - 获取翻译详情

获取翻译任务的详细信息和状态。

参数:

  • task_id (必填): 翻译任务 ID

示例:

{
  "task_id": "02a6c6d1-3f70-4a5a-80bc-971d53a37bb1"
}

返回:

{
  "task_id": "02a6c6d1-3f70-4a5a-80bc-971d53a37bb1",
  "status": "DONE",
  "progress": 1.0,
  "consumed_point": 453,
  "source_file_name": "document.pdf",
  "source_file_url": "https://example.com/source.pdf",
  "target_file_url": "https://example.com/translated.pdf",
  "source_lang": "zh",
  "target_lang": "en",
  "error_msg": null,
  "optimize_math_formula": true
}

任务状态说明:

  • INIT: 初始化
  • PROGRESS: 进行中
  • DONE: 已完成
  • ERROR: 错误
3. list_translations - 列出翻译任务

获取翻译任务列表,支持分页。

参数:

  • offset (可选): 分页偏移量,默认 0
  • limit (可选): 每页数量,默认 20

示例:

{
  "offset": 0,
  "limit": 10
}

返回:

{
  "total": 42,
  "offset": 0,
  "limit": 10,
  "list": [
    {
      "task_id": "...",
      "status": "DONE",
      "progress": 1.0,
      ...
    }
  ]
}
4. search_documents - 文献搜索

AI 驱动的文献语义搜索。

参数:

  • query (必填): 自然语言查询
  • topk (可选): 最大返回数量(1-100,默认 20)
  • return_doc_keys (可选): 指定返回字段
  • auto_select (可选): 自动选择最优结果(默认 true)

示例:

{
  "query": "糖尿病最新研究进展",
  "topk": 5,
  "return_doc_keys": ["title", "abstract", "doi", "authors"],
  "auto_select": true
}

可用的返回字段:

  • title: 标题
  • abstract: 摘要
  • authors: 作者列表
  • doi: DOI
  • pmid: PubMed ID
  • link: 链接
  • publication: 出版物
  • pub_year: 出版年份
  • 更多字段请参考 API 文档

返回:

{
  "search_items": [
    {
      "doc": {
        "title": "...",
        "abstract": "...",
        "authors": [...],
        "doi": "...",
        ...
      },
      "search_gateway": "pubmed"
    }
  ],
  "consumed_points": 20
}

支持的语言

常用语言代码:

  • en: English (英语)
  • zh: Chinese (中文)
  • ko: Korean (韩语)
  • ja: Japanese (日语)
  • fr: French (法语)
  • de: German (德语)
  • es: Spanish (西班牙语)
  • ru: Russian (俄语)
  • ar: Arabic (阿拉伯语)
  • pt: Portuguese (葡萄牙语)
  • it: Italian (意大利语)
  • auto: 自动检测

错误处理

所有错误都会返回标准格式:

{
  "code": 非零错误码,
  "msg": "错误信息",
  "data": null
}

常见错误:

  • 401: API 密钥无效或未提供
  • 400: 请求参数错误
  • 404: 资源不存在

使用示例

在 Claude Desktop 中使用
  1. 配置好 API 密钥后重启 Claude Desktop

  2. 在对话中使用工具:

翻译文档:

请帮我翻译这个文档:https://example.com/paper.pdf,翻译成英文

搜索文献:

帮我搜索关于"深度学习在医学影像中的应用"的最新文献

查询翻译状态:

查看任务 02a6c6d1-3f70-4a5a-80bc-971d53a37bb1 的翻译进度

常见问题

Q: 如何获取 API 密钥?

A: 访问 https://suppr.wilddata.cn/api-keys 注册并获取 API 密钥。

Q: 支持哪些文档格式?

A: 支持 PDF, DOCX, PPTX, XLSX, HTML, TXT, EPUB等常见格式。

Q: 翻译需要多长时间?

A: 取决于文档大小,通常几分钟到十几分钟不等。可以使用 get_translation 查询进度。

Q: 如何下载翻译后的文档?

A: 翻译完成后,get_translation 会返回 target_file_url,直接访问该链接下载。

Q: npx 运行失败?

A: 确保 Node.js 版本 >= 18.0.0,并且设置了 SUPPR_API_KEY 环境变量。

🔗 Suppr超能文献产品

技术支持

如需帮助,请联系:[email protected]

Made with ❤️ by WildData

Suppr Ecosystem

ProductLink
🌐 Suppr Platformsuppr.wilddata.cn
📖 API Documentationopenapi.suppr.wilddata.cn
🔌 Zotero Pluginsuppr-zotero-plugin
🤖 Claude Code Skillssuppr-skills
🔬 Deep Researchsuppr.wilddata.cn/deep-research
📄 AI Translationsuppr.wilddata.cn/translate/upload
🔎 PubMed Searchsuppr.wilddata.cn

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.