Mcp Pdf Reader

What is Mcp Pdf Reader about?

Mcp Pdf Reader provides a Model Context Protocol (MCP) server that enables applications to extract raw text, perform advanced searches, and pull detailed metadata from PDF documents. It is designed for developers who need reliable PDF processing within chat‑based or automated workflows.

How to use Mcp Pdf Reader?

1. Installation – Clone the repository and run npm install to resolve dependencies.
2. Running the server – Start the MCP server (e.g., node index.js in the project root) and configure your MCP client to point to the server executable.
3. Invoke tools – Use the three provided tools via JSON payloads:
   • read-pdf for text extraction.
   • search-pdf for searching specific terms.
   • pdf-metadata for retrieving only metadata. Each tool accepts parameters such as file path, page ranges, case‑sensitivity, and cleaning options.

Key features of Mcp Pdf Reader

• Text Extraction with optional page ranges, metadata inclusion, and text cleaning.
• Smart Search supporting case‑sensitive, whole‑word, and regex‑like queries.
• Rich Metadata extraction covering author, title, dates, keywords, encryption status, and more.
• Async Processing ensures non‑blocking operations.
• File Size Guardrails (50 MB limit) to prevent resource exhaustion.
• Security through file validation and path sanitization.
• Three dedicated tools for focused functionality.

Use cases of Mcp Pdf Reader

• Integrating PDF text search into a knowledge‑base chatbot.
• Automating extraction of contract details (authors, dates, keywords) for compliance pipelines.
• Building document summarization or indexing services that need raw text from large PDF collections.
• Enabling developers to quickly prototype PDF‑driven workflows without writing low‑level parsing code.

FAQ from the Mcp Pdf Reader

Q: What is the maximum PDF size supported? A: Files larger than 50 MB are rejected to protect the server from excessive memory use.

Q: Can I extract text from scanned or image‑based PDFs? A: Not yet. OCR support is planned for a future release.

Q: How do I include or exclude metadata during extraction? A: Set the include_metadata boolean flag in the read-pdf request payload (default is true).

Q: Is there support for password‑protected PDFs? A: Password handling is on the roadmap; currently only unprotected PDFs are processed.

Q: How can I run the server from a different directory? A: Provide the absolute path to the index.js entry point in your MCP client configuration, as shown in the README example.

MCP PDF Reader Enhanced

A comprehensive Model Context Protocol (MCP) server that provides advanced PDF text extraction, search, and analysis functionality.

Features

Core Functionality

• ✅ Text Extraction: Extract text content from PDF files with customizable options
• ✅ Text Search: Search for specific text within PDFs with advanced options
• ✅ Metadata Extraction: Retrieve comprehensive PDF metadata
• ✅ Page-specific Processing: Extract content from specific page ranges
• ✅ Text Cleaning: Normalize and clean extracted text
• ✅ File Size Limits: Protection against overly large files (50MB limit)
• ✅ Async Processing: Non-blocking file operations

Advanced Features

• 🔄 Multiple Tools: 3 specialized tools for different PDF operations
• 🔍 Smart Search: Case-sensitive, whole-word, and regex search options
• 📊 Rich Metadata: Extract author, title, creation date, keywords, and more
• ⚡ Performance: Efficient processing with size limits and error handling
• 🛡️ Security: File validation and path sanitization

Installation

npm install

Tools Available

1. read-pdf - Enhanced PDF Reading

Extract text from PDF files with customizable options.

Parameters:

• file (string, required): Path to the PDF file
• pages (string, optional): Page range (e.g., '1-5', '1,3,5', 'all'). Default: 'all'
• include_metadata (boolean, optional): Include PDF metadata. Default: true
• clean_text (boolean, optional): Clean and normalize text. Default: false

Example Usage:

// Basic extraction
{ "file": "/path/to/document.pdf" }

// Extract with clean text and no metadata
{ 
  "file": "/path/to/document.pdf", 
  "clean_text": true, 
  "include_metadata": false 
}

2. search-pdf - Search Within PDFs

Search for specific text within PDF documents.

Parameters:

• file (string, required): Path to the PDF file
• query (string, required): Text to search for
• case_sensitive (boolean, optional): Case sensitive search. Default: false
• whole_word (boolean, optional): Match whole words only. Default: false

Example Usage:

// Case-insensitive search
{ "file": "/path/to/document.pdf", "query": "important term" }

// Whole word, case-sensitive search
{ 
  "file": "/path/to/document.pdf", 
  "query": "API", 
  "case_sensitive": true, 
  "whole_word": true 
}

3. pdf-metadata - Extract Metadata Only

Get comprehensive metadata from PDF files without extracting text.

Parameters:

• file (string, required): Path to the PDF file

Returns:

• Filename, file size, page count
• Author, title, subject, creator, producer
• Creation/modification dates, keywords
• Encryption status, PDF version

Configuration

Cursor Integration

Add to your Cursor settings:

{
  "mcpServers": {
    "mcp-gp-pdf-reader": {
      "command": "node",
      "args": ["/absolute/path/to/mcp_gp_pdf_reader/index.js"]
    }
  }
}

Future Enhancements

Planned Features

• 🔮 OCR Support: Extract text from scanned/image-based PDFs
• 🔮 Image Extraction: Extract images from PDF documents
• 🔮 Table Detection: Identify and extract tabular data
• 🔮 Form Data: Extract form fields and values
• 🔮 Password Support: Handle password-protected PDFs
• 🔮 Batch Processing: Process multiple PDFs simultaneously
• 🔮 Caching: Cache parsed results for better performance
• 🔮 Page-by-Page: True page-specific text extraction

Technical Improvements

• 🔧 Streaming: Handle very large PDFs with streaming
• 🔧 Progress Tracking: Progress indicators for long operations
• 🔧 Resource Management: Better memory usage optimization
• 🔧 Configuration API: Runtime configuration updates

Usage Examples

Basic Text Extraction

# Via MCP client
"Extract all text from /documents/report.pdf"

Searching PDFs

# Via MCP client  
"Search for 'quarterly results' in /documents/financial-report.pdf"

Getting Metadata

# Via MCP client
"Get metadata from /documents/contract.pdf"

Development

Requirements

• Node.js 18.0.0 or higher
• Memory: Sufficient for PDF file size + processing overhead
• Storage: Temporary space for file operations

Contributing

This MCP server is designed to be extensible. Key areas for contribution:

• Additional PDF processing libraries integration
• Performance optimizations
• New extraction features
• Better error handling
• Test coverage

License

MIT License