book-to-skill

book-to-skill is an open-source skill from GitHub user virgiliojr94. In the README's words, it will "Turn any technical book, document folder, or collection of sources into a unified agent skill." It first extracts the text locally from PDF, EPUB, DOCX, HTML, MOBI, Markdown and other formats, then has your AI assistant organize it into a set of notes: a core-concepts file with a chapter index, one file per chapter, plus a glossary, a patterns file and a cheat sheet. When you ask a question later, the AI loads only the relevant chapter instead of the whole book. Besides books, it can handle internal docs, brand guidelines, a cluster of research papers or technical specs. It works with Claude Code, GitHub Copilot CLI, Codex, OpenCode and other tools that support the Agent Skills standard.

Who it's for:Learners and teams with a shelf of technical books or documents who want their AI assistant to cite them chapter by chapter.

Before you start: Install with npx skills add virgiliojr94/book-to-skill, or clone the repository into your assistant's skills folder; extracting different formats needs Python 3 and the matching libraries, and scanned PDFs need OCR first. Generated skills go to ~/.agents/skills by default, and under Claude Code it also creates a link, which changes your assistant's skill setup. Extraction runs locally, but anything sent to a cloud model follows that provider's terms. There's an optional publish-to-GitHub step that defaults to private and asks for your consent first. The MIT license covers the tool only; books and documents stay with their rights holders, so only process material you have the right to use and keep the generated notes to yourself. The author warns that a malicious re-upload with the same name exists elsewhere and that this repository is the only official source. The README claims large token savings, which we haven't verified.

Based on the project's public documentation. Not tested by AI Inspiration World.