Releases

Version history and what shipped

From the first stable multi format converter to an installable package. Follow the changelog and jump to each GitHub release tag.

View source on GitHub

Latest: installable package

Starting with v1.2.0, the extractor is a proper book_to_skill package: pyproject.toml, a book-to-skill console script, and python -m book_to_skill. Optional extractors ship as extras (epub, pdf, docx, rtf, technical, all). The agent skill still installs via git clone into your skills folder.

Release notes

v1.3.02026-07-30
Korean and Thai chapter detection, prompt injection scan for generated skills, and extraction hardening (invisible Unicode, DOCX XXE, subprocess args).
View on GitHub ↗
v1.2.02026-06-17
Turns the project into an installable Python package with a book-to-skill console script, optional format extras, and broader multilingual heading detection.
View on GitHub ↗
v1.1.02026-06-12
GitHub Copilot CLI as a first class target via the open Agent Skills standard, plus validate_skill.py lenses for claude, copilot, and amp.
View on GitHub ↗
v1.0.02026-06-08
First tagged release with multi format extraction, adaptive chapter depth, discovery tax tooling, update and fold in workflow, and CI.
View on GitHub ↗

Install the latest

# Agent skill (slash command)
git clone https://github.com/virgiliojr94/book-to-skill.git ~/.claude/skills/book-to-skill

# Standalone CLI only
pip install "book-to-skill[pdf,epub,docx]"

Full changelog: CHANGELOG.md on GitHub.

Review security practices

The tool is local first. See how vulnerabilities are reported and what users should verify before converting files.

Open SECURITY.md