Suppr MCP

🚀 Suppr MCP - Usage Guide | Provide document translation and MCP service for Chinese PubMed search | Suppr Super Literature

This project offers a comprehensive MCP service, enabling document translation and Chinese PubMed search. It provides a set of useful tools to meet various user needs.

🚀 Quick Start

1. Installation

You can choose to install the tool globally or use it with npx without installation.

• Global Installation:

npm install -g suppr-mcp

• Using npx:

npx suppr-mcp

2. Obtain API Key

Visit Suppr API to get your API key.

3. Configure Environment Variables

export SUPPR_API_KEY=your_api_key_here

4. Use in MCP Client

Claude Desktop Configuration

Edit ~/Library/Application Support/Claude/claude_desktop_config.json (for macOS) or the corresponding configuration file:

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

Or use the globally installed version:

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

✨ Features

1. create_translation - Create a Translation Task

Create a document translation task.

Parameters:

• file_path (either file_path or file_url is required): Path to the source file.
• file_url (either file_path or file_url is required): URL of the document to be translated.
• to_lang (required): Target language code.
• from_lang (optional): Source language code (automatically detected by default).
• optimize_math_formula (optional): Optimize mathematical formulas (only for PDFs).

Example:

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

Return:

{
  "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 - Get Translation Details

Get detailed information and status of a translation task.

Parameters:

• task_id (required): Translation task ID.

Example:

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

Return:

{
  "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
}

Task Status Explanation:

• INIT: Initialization
• PROGRESS: In progress
• DONE: Completed
• ERROR: Error

3. list_translations - List Translation Tasks

Get a list of translation tasks with pagination support.

Parameters:

• offset (optional): Pagination offset, default is 0.
• limit (optional): Number of items per page, default is 20.

Example:

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

Return:

{
  "total": 42,
  "offset": 0,
  "limit": 10,
  "list": [
    {
      "task_id": "...",
      "status": "DONE",
      "progress": 1.0,
      ...
    }
  ]
}

4. search_documents - Document Search

AI-driven semantic search for documents.

Parameters:

• query (required): Natural language query.
• topk (optional): Maximum number of returned results (1 - 100, default is 20).
• return_doc_keys (optional): Specify the fields to return.
• auto_select (optional): Automatically select the best result (default is true).

Example:

{
  "query": "Latest research progress on diabetes",
  "topk": 5,
  "return_doc_keys": ["title", "abstract", "doi", "authors"],
  "auto_select": true
}

Available Return Fields:

• title: Title
• abstract: Abstract
• authors: List of authors
• doi: DOI
• pmid: PubMed ID
• link: Link
• publication: Publication
• pub_year: Publication year
• For more fields, refer to the API documentation.

Return:

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

📚 Supported Languages

Common language codes:

• en: English
• zh: Chinese
• ko: Korean
• ja: Japanese
• fr: French
• de: German
• es: Spanish
• ru: Russian
• ar: Arabic
• pt: Portuguese
• it: Italian
• auto: Automatic detection

🔧 Error Handling

All errors will be returned in a standard format:

{
  "code": non-zero error code,
  "msg": "Error message",
  "data": null
}

Common errors:

• 401: Invalid or missing API key.
• 400: Invalid request parameters.
• 404: Resource not found.

💻 Usage Examples

Usage in Claude Desktop

1. Restart Claude Desktop after configuring the API key.
2. Use the tools in the conversation:

Translate a Document:

Please translate this document: https://example.com/paper.pdf into English.

Search for Documents:

Search for the latest research on "applications of deep learning in medical imaging".

Check Translation Status:

Check the translation progress of task 02a6c6d1-3f70-4a5a-80bc-971d53a37bb1.

📋 Frequently Asked Questions

Q: How can I get an API key?

A: Visit https://suppr.wilddata.cn/api-keys to register and obtain an API key.

Q: What document formats are supported?

A: Common formats such as PDF, DOCX, PPTX, XLSX, HTML, TXT, EPUB are supported.

Q: How long does translation take?

A: It depends on the document size, usually ranging from a few minutes to tens of minutes. You can use get_translation to check the progress.

Q: How can I download the translated document?

A: After the translation is completed, get_translation will return the target_file_url. Simply access this link to download the document.

Q: Why does npx fail to run?

A: Make sure your Node.js version is >= 18.0.0 and the SUPPR_API_KEY environment variable is set.

🔗 Suppr Super Literature Products

• Zotero Plugin: https://github.com/WildDataX/suppr-zotero-plugin
• Official Website: https://suppr.wilddata.cn
• AI Document Translation: https://suppr.wilddata.cn/translate/upload
• API Service: https://openapi.suppr.wilddata.cn/introduction
• Chinese PubMed Search: https://suppr.wilddata.cn/
• In-depth Research: https://suppr.wilddata.cn/deep-research
• GitHub Organization: WildDataX

📞 Technical Support

If you need assistance, please contact: IT@wilddata.cn

Made with ❤️ by WildData