you buy a great technical book. you read it once. three months later you can't remember what chapter 7 was even about.
the usual workarounds don't work:
- "let me search the PDF" → you get page numbers, not answers
- "i'll ask Claude about this book" → it either hallucinates or says it doesn't have the content
- "i'll take notes as I read" → you end up with a 200-line doc you never open again
so I built /book-to-skill.
you give it a PDF path. it extracts the text, analyzes the chapter structure, generates dense chapter summaries (800–1200 tokens each, loaded on-demand so they don't bloat your context), a glossary, a patterns file, and a master SKILL.md with the author's core frameworks.
after that, the book becomes a skill you just use:
/designing-data-intensive-apps replication ← explains replication from the actual book
/clean-code ch07 ← dives into chapter 7
/cialdini-influence scarcity principle ← applies the framework to what you're working on
it's not a summary. the goal is to extract structure — named frameworks, exact formulations, anti-patterns, mental models. the kind of stuff that took the author years to crystallize. once it's a skill, Claude can reason with it while you work, not just recite it.
the skill itself handles everything: validates the PDF, runs extraction with three fallbacks (pdftotext → PyPDF2 → pdfminer), gives you a token cost estimate before doing anything, and asks what you want to use the skill for before generating.
install (one command in any Claude Code session):
Install book-to-skill: https://raw.githubusercontent.com/virgiliojr94/book-to-skill/master/SKILL.md
then:
/book-to-skill ~/Downloads/your-book.pdf
repo: https://github.com/virgiliojr94/book-to-skill
caveats:
- works best on digital PDFs (scanned/image PDFs need OCR first)
- heavy skill — takes a few minutes on a long book
- each chapter file is on-demand so the base skill stays under 4K tokens