A comprehensive set of PDF manipulation tools built with the Model Context Protocol (MCP) framework.
https://github.com/danielkennedy1/pdf-tools-mcpYou've been there: juggling PyPDF2, pdfplumber, and custom scripts just to extract text from a research paper or merge a few invoices. Then you need to handle remote PDFs, and suddenly you're writing download logic too.
PDF Tools MCP puts an end to that workflow chaos.
Every developer has a folder full of half-finished PDF scripts. Extract text from this document, merge these pages, grab metadata from that file. Each PDF task becomes a mini-project with different libraries, error handling, and file management headaches.
When you're working with AI models that need to process PDFs, the friction multiplies. You're constantly context-switching between your AI workflow and manual PDF preprocessing.
This MCP server gives you comprehensive PDF manipulation directly in your AI workflow. No more preprocessing scripts or separate tools.
Local PDF Operations:
Remote PDF Operations:
Research Workflow: Instead of downloading papers, converting them, and manually extracting sections, ask Claude to analyze specific pages from remote research PDFs directly through URLs.
Document Processing: Need to merge invoice pages from different months? Skip the manual PDF editor - do it programmatically through your AI workflow.
Content Analysis: Extract text blocks with precise positioning data for layout-aware document analysis, not just raw text dumps.
Report Generation: Combine specific pages from multiple reports into a single document, then immediately analyze the consolidated content.
Uses PyMuPDF for robust PDF handling with async operations throughout. Documents get UUID4 filenames for security, and the async architecture means you're not blocking on large PDF operations.
The MCP protocol integration means this works seamlessly with Claude Desktop and any other MCP-compatible tools you're already using.
Install through Smithery for Claude Desktop:
npx -y @smithery/cli install @danielkennedy1/pdf-tools-mcp --client claude
Or clone and run locally:
git clone https://github.com/danielkennedy1/pdf-tools-mcp.git
cd pdf-tools-mcp
uv pip install -e .
python -m src.main
Your PDF workflows just got significantly simpler. Instead of maintaining a collection of scripts and switching between tools, you now have comprehensive PDF capabilities directly in your AI workflow.
Time to delete those half-finished PDF scripts cluttering your projects folder.