KnowledgeBaseMCP

A powerful Model Context Protocol (MCP) server for extracting text content from various document formats including PDF, DOCX, PPTX, and XLSX files. This tool enables AI assistants like Claude to read and analyze document contents from your local knowledge base, and also create new Excel spreadsheets.

🚀 Features

Document Reading

• Multi-format support: Extract text from PDF, DOCX, PPTX, and XLSX files

• Directory processing: Process entire directories of documents

• Recursive scanning: Optionally scan subdirectories

• File metadata: Get detailed information about document files

• Error handling: Robust error handling with clear error messages

• Async processing: Efficient asynchronous document processing

Excel Spreadsheet Creation

• XLSX workbook creation: Create Excel files with multiple sheets

• DataFrame support: Convert pandas DataFrames to Excel

• Data formatting: Apply professional formatting and styling

• Report generation: Create structured reports with summaries

• Data appending: Add data to existing Excel files

• Template support: Use predefined templates for consistent formatting

Integration

• Easy integration: Simple setup with Claude Desktop

• MCP protocol: Built on the Model Context Protocol standard

📁 Supported File Types

Reading Support

• PDF (.pdf) - Portable Document Format (using pdfplumber)

• DOCX (.docx) - Microsoft Word documents

• PPTX (.pptx) - Microsoft PowerPoint presentations

• XLSX (.xlsx) - Microsoft Excel spreadsheets (using openpyxl and pandas)

Writing Support

• DOCX (.docx) - Create Word documents with formatting

• XLSX (.xlsx) - Create Excel workbooks with multiple sheets, formatting, and charts

🛠️ Installation

Prerequisites

• Python 3.8 or higher

• Claude Desktop application

Setup

1. Clone the repository

2. Install dependencies

3. Test the server

⚙️ Configuration

Claude Desktop Integration

Add this server to your Claude Desktop configuration file:

Windows: %APPDATA%\\Claude\\claude_desktop_config.json
macOS: ~/Library/Application Support/Claude/claude_desktop_config.json

Replace path/to/KnowledgeBaseMCP with your actual installation path.

🎯 Usage

Once configured, you can use these tools in Claude:

Available Tools

Document Reading Tools

extract_text_from_file

Extract text content from a single document file.

Parameters:

• file_path (string): Path to the document file

extract_text_from_directory

Extract text content from all supported documents in a directory.

Parameters:

• directory_path (string): Path to the directory containing documents

• recursive (boolean, optional): Whether to search subdirectories recursively

list_supported_files

List all supported document files in a directory with metadata.

Parameters:

• directory_path (string): Path to the directory to scan

DOCX Creation Tools

create_docx_document

Create a new Word document with text content.

Parameters:

• content (string): Document content

• file_path (string): Output file path (.docx extension)

• title (string, optional): Document title

create_structured_report

Create a structured Word report with formatting.

Parameters:

• report_data (object): Report data structure

• file_path (string): Output file path (.docx extension)

XLSX Creation Tools

create_xlsx_workbook

Create a new Excel workbook with multiple sheets.

Parameters:

• data (object): Dictionary with sheet names as keys and data as values

• file_path (string): Output file path (.xlsx extension)

• apply_formatting (boolean, optional): Apply default formatting

create_xlsx_from_dataframe

Create Excel workbook from pandas DataFrames.

Parameters:

• dataframes (object): Dictionary with sheet names and DataFrame data

• file_path (string): Output file path (.xlsx extension)

• include_index (boolean, optional): Include DataFrame index

append_to_xlsx

Append data to existing Excel workbook.

Parameters:

• file_path (string): Path to existing XLSX file

• sheet_name (string): Target sheet name

• data (any): Data to append (list, dict, or DataFrame)

create_xlsx_report

Create a formatted Excel report with multiple sections.

Parameters:

• report_data (object): Report structure with title, description, and data sections

• file_path (string): Output file path (.xlsx extension)

Example Usage in Claude

Reading Documents

Creating Excel Reports

Data Analysis and Export

Document Conversion

Claude will then use the MCP server to extract and analyze the content from your documents or create new Excel files as requested.

🏗️ Project Structure

🔧 Development

Running Tests

Adding New File Formats

To add support for additional document formats:

1. Add the file extension to SUPPORTED_EXTENSIONS in extractors.py

2. Install the required library

3. Add the library check to check_dependencies()

4. Implement the extraction method (e.g., _extract_xlsx())

5. Add the format handling to extract_from_file()

Debugging

For debugging MCP connection issues:

1. Check Claude Desktop logs

2. Ensure the server starts without errors:

   python launch_mcp.py

3. Verify the config file path and format

📦 Dependencies

Core Dependencies

• mcp>=0.9.0 - Model Context Protocol framework

Document Reading

• python-docx>=1.1.0 - For DOCX file processing

• pdfplumber>=0.9.0 - For PDF file processing

• python-pptx>=0.6.23 - For PPTX file processing

• openpyxl>=3.1.0 - For XLSX file reading/writing

• pandas>=2.0.0 - For advanced data manipulation and analysis

Additional

• lxml>=4.9.0 - XML processing support

🤝 Contributing

Contributions are welcome! Please feel free to submit pull requests or open issues.

1. Fork the repository

2. Create your feature branch (git checkout -b feature/AmazingFeature)

3. Commit your changes (git commit -m 'Add some AmazingFeature')

4. Push to the branch (git push origin feature/AmazingFeature)

5. Open a Pull Request

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

🙏 Acknowledgments

• Built with the Model Context Protocol

• Uses pdfplumber for PDF processing

• Uses python-docx for Word documents

• Uses python-pptx for PowerPoint presentations

📞 Support

If you encounter any issues or have questions, please open an issue on GitHub.

Made with ❤️ for the Claude AI community