Skill

pdf-to-ebook

Converts PDF files to EPUB, MOBI, or AZW3 ebook formats using Calibre's ebook-convert via Bash. Validates files, detects scanned PDFs for optional OCR with ocrmypdf, and lists outputs. Use for PDF-to-ebook tasks.

Bash

cli-tools

automation

npx claudepluginhub cgbarlow/skills --plugin math-coach

Tool Access

This skill is limited to using the following tools:

Bash

Preview

Convert a PDF to one or more ebook formats using Calibre's `ebook-convert`.

Supporting Assets

README.md

SKILL.md

Similar Skills

translate-book-parallel

Translates entire books (PDF/DOCX/EPUB) to any language using Claude Code parallel subagents, resumable chunked Markdown pipeline, Calibre conversion, and Pandoc merging.

aradotso-trending-skills-37

document-converter

Converts Office docs (DOCX/PPTX/XLSX) to PDF/HTML, PDFs to images/images to PDF, performs PDF merge/split/rotate/encrypt/decrypt, and OCR on scans using local tools: LibreOffice, Ghostscript, pdftk, Tesseract, ImageMagick.

1 file

claude-superskills

pdf-generation

Generates styled PDF books from directories of ordered markdown chapters using pandoc and weasyprint, with TOC, page numbers, and resolved inter-chapter links.

2 files

technical-writer

Stats

Stars17

Forks8

Last CommitMay 4, 2026

Actions

View Source View Plugin View on GitHub View README

Help us improve

Share bugs, ideas, or general feedback.

PDF to Ebook Converter

Convert a PDF to one or more ebook formats using Calibre's ebook-convert.

Arguments

$ARGUMENTS — space-separated: first arg is the PDF path, optional second arg is the format(s).

Format options: epub, mobi, azw3, all
If no format is provided, ask the user which formats they want before proceeding.

Workflow

Step 1 — Parse arguments

Extract the PDF path from $ARGUMENTS. If a format argument is also present, use it. If not, ask:

Which formats would you like to export?

epub

mobi

azw3 (Kindle, recommended over mobi)

all (epub + mobi + azw3)

Wait for the user's response before continuing.

Step 2 — Validate inputs

Confirm the PDF file exists. If not, report the error and stop.
Check Calibre is installed:
```
which ebook-convert
```
If missing, tell the user to install it:
- Linux: sudo apt install calibre
- macOS: brew install calibre
- Then stop.

Step 3 — Check for scanned/image-only PDF (optional but recommended)

Run a quick text extraction check:

ebook-convert "<input.pdf>" /tmp/pdf_text_check.txt 2>/dev/null && wc -c /tmp/pdf_text_check.txt

If the output is fewer than 500 characters, the PDF is likely image-based (scanned). Warn the user:

This PDF appears to be scanned or image-based and may convert poorly. Would you like to run OCR first using ocrmypdf? (Requires: pip install ocrmypdf and tesseract)

If they say yes:

ocrmypdf "<input.pdf>" "<input_ocr.pdf>"

Use the OCR'd file for all subsequent conversions.

Step 4 — Convert

Derive the output base name from the input file (strip .pdf extension).

For each requested format, run:

ebook-convert "<input.pdf>" "<basename>.<format>" \
  --output-profile kindle_oasis \
  --enable-heuristics

Format-specific notes:

epub: drop --output-profile; use --output-profile tablet instead
mobi: use --output-profile kindle
azw3: use --output-profile kindle_oasis (best Kindle compatibility)

Run conversions sequentially. Report each one as it completes.

Step 5 — Report results

List each output file with its path and size:

ls -lh "<basename>.epub" "<basename>.mobi" "<basename>.azw3" 2>/dev/null

If any conversion failed, show the error output from ebook-convert.

Example invocations

/pdf-to-ebook book.pdf
/pdf-to-ebook book.pdf epub
/pdf-to-ebook book.pdf all
/pdf-to-ebook /path/to/document.pdf azw3