book-to-skill
book-to-skill is an open-source tool and agent skill by Virgilio Jr that converts technical books, document folders, and API specs into unified, structured Agent Skills. Rather than dumping an entire multi-hundred-page PDF into a model's context window—which subjects every subsequent query to repetitive 'discovery loop taxes' as the model refetches tables of contents and navigates sections—the tool distills the source material once into a persistent, modular knowledge bundle.
The resulting artifact adheres to the open Agent Skills standard. It generates a lightweight `SKILL.md` (around 4,000 tokens) containing high-level mental models and chapter indexes, complemented by separate files for chapter deep-dives (`chapters/ch01-*.md`), an alphabetized `glossary.md`, core architectural `patterns.md`, and decision-tree `cheatsheet.md`. Crucially, detailed chapter files are loaded by the agent on demand, enabling 24× to 51× token savings on individual technical queries while eliminating context-drift hallucinations.
It supports PDF, EPUB, DOCX, Markdown, and MOBI sources, installing globally via `npx skills add virgiliojr94/book-to-skill` for GitHub Copilot CLI, Claude Code, Codex, and OpenClaw. Users should be prepared for significant upfront time and LLM token expenditure during the initial conversion pass of lengthy books. Additionally, noisy scanned documents with non-standard multi-column layouts or complex figures can suffer degraded extraction quality.
Features
Why choose it
Trade-offs
Maintenance check
- Last commit
- 2026-11-03
- Commits, last 90 days
- 100+
- Releases, last 90 days
- 0
- Latest release
- v1.4.0 (2026-08-10)
- Open issues
- 27
- Recent issues with at least one comment
- 2/13
Oldest week on the left; the latest four weeks are orange. Source: GitHub commit activity.
Status rule: archived on GitHub = Archived; last commit within 90 days = Active; 91 to 365 days = Slowing; older = Stale. All figures come from the GitHub API on the date shown.
Most discussed open issues
- #309 Proposal: slim down the installed skill payload (split runtime files from dev/site files)4 comments · 2026-10-27
- #169 The generator's own SKILL.md is 7k tokens with zero references — 52% of it loads before the run knows it can proceed2 comments · 2026-08-15
- #128 PDF: `pdftotext -layout` interleaves columns on multi-column documents, and the word-based token estimate hides the 26x size blowup3 comments · 2026-08-10
- #263 [bug] BeautifulSoup HTML/EPUB extraction inserts newlines inside inline headings and code tokens1 comments · 2026-10-01
- #259 [bug] Interrupted-run reuse accepts missing or corrupted extracted text1 comments · 2026-09-30
Issue titles are quoted from GitHub. Open the link to check whether a problem still applies to the version you use.
Compatibility
Quick start
Use cases
Alternatives
Comments
No comments yet. Be the first!