ebook-extractor

by @ratacat in Tools

# Install this skill:

npx skills add ratacat/claude-skills --skill "ebook-extractor"

Install specific skill from multi-skill repository

# Description

Use when user wants to extract text from ebooks (EPUB, MOBI, PDF). Use for converting ebooks to plain text for analysis, processing, or reading. Handles all common ebook formats.

# SKILL.md

name: ebook-extractor
description: Use when user wants to extract text from ebooks (EPUB, MOBI, PDF). Use for converting ebooks to plain text for analysis, processing, or reading. Handles all common ebook formats.

Ebook Text Extractor

Overview

Extract plain text from EPUB, MOBI, and PDF files using Python scripts. No LLM calls - pure text extraction.

Supported Formats

Format	Tool Used	Notes
EPUB	`ebooklib` + `BeautifulSoup`	Direct parsing, preserves structure
MOBI	Calibre `ebook-convert`	Converts to EPUB first, then extracts
PDF	`PyMuPDF` (fitz)	Fast, handles most PDFs well

Usage

Unified extractor (auto-detects format):

python3 ~/.claude/skills/ebook-extractor/scripts/extract.py /path/to/book.epub
python3 ~/.claude/skills/ebook-extractor/scripts/extract.py /path/to/book.mobi
python3 ~/.claude/skills/ebook-extractor/scripts/extract.py /path/to/book.pdf

Output options:

# To stdout (default)
python3 scripts/extract.py book.epub

# To file
python3 scripts/extract.py book.epub -o output.txt
python3 scripts/extract.py book.epub > output.txt

Format-specific scripts:

python3 scripts/extract_epub.py book.epub
python3 scripts/extract_mobi.py book.mobi
python3 scripts/extract_pdf.py book.pdf

Setup

# One-command setup (installs all dependencies)
~/.claude/skills/ebook-extractor/setup.sh

# Or manually:
pip install -r ~/.claude/skills/ebook-extractor/requirements.txt
brew install calibre  # macOS, for MOBI support

Script Location

~/.claude/skills/ebook-extractor/scripts/

Common Issues

Problem	Solution
Missing package	Run `setup.sh` or `pip install -r requirements.txt`
MOBI fails	Ensure Calibre is installed: `brew install calibre`
PDF garbled	Some PDFs are image-based; OCR needed (not supported)