PDF Reader MCP Server

A Model Context Protocol (MCP) server for reading and searching PDF files.

Installation

Option 1: Quick Install

pip install git+https://github.com/bencode/pdf-reader-mcp.git@stable
claude mcp add pdf-reader pdf-reader-mcp

Option 2: Development Setup

git clone https://github.com/bencode/pdf-reader-mcp.git
cd pdf-reader-mcp
./install.sh

Testing

# Run all tests
pytest tests/ -v

# Run tests with coverage
pytest tests/ --cov=pdf_reader_server

# Using uv
uv run pytest tests/ -v

Usage

1. Development Mode

# Using pip installation
mcp dev pdf_reader_server.py

# Or using the installed script
pdf-reader-mcp

2. Install to Claude Code

Based on the Claude Code MCP documentation:

# For Option 1 (quick install) - already configured above ✅

# For Option 2 (development setup) - replace /absolute/path with your actual path
claude mcp add pdf-reader /absolute/path/increa-pdf-reader/.venv/bin/python /absolute/path/increa-pdf-reader/pdf_reader_server.py

3. Available Tools

The server provides these tools that Claude can use:

1. open_pdf(path) - Open PDF and return doc_id
2. page_count(doc_id) - Get number of pages
3. extract_text(doc_id, page) - Extract text from page
4. render_page_png(doc_id, page, dpi) - Render page as PNG
5. search_text(doc_id, query, max_hits) - Search for text
6. close_pdf(doc_id) - Close document