Book to skill
virgiliojr94/book-to-skill
Book to skill is a free, MIT-licensed agent skill by virgiliojr94 that converts a book or document into a structured skill. It reads PDF, EPUB, DOCX, Markdown, HTML, RTF and MOBI files, then writes a skill with a core summary, one file per chapter, a glossary, patterns and a cheatsheet that the agent loads on demand.
Facts checked on against the sources listed at the end of this page.
- GitHub stars
- 34,308
- Skills in the pack
- 1
- Agents supported
- 6plus any agent the skills CLI supports
- License
- MIT
- Last commit
- 3 days ago5 October 2026
- Instruction size
- 7.5k wordsacross its SKILL.md file
How to install Book to skill
| Agent | Command | Copy |
|---|---|---|
| Any agent (skills CLI) | | |
| Claude Code | Registers the /book-to-skill command. | |
| GitHub Copilot CLI | Place the skill folder in ~/.copilot/skills/. | |
| Amp and other cross-agent setups | Place the skill folder in ~/.agents/skills/. OpenCode reads the same folder. | |
| Hermes Agent | Place the skill folder in ${HERMES_HOME:-$HOME/.hermes}/skills/<category>/. |
Commands for 1 more agent
| OpenClaw | Place the skill folder in ${OPENCLAW_STATE_DIR:-$HOME/.openclaw}/skills/. |
|---|
Commands as given in the README on 9 October 2026.
What Book to skill runs on your machine
- Hooks
- None named in the README.
- Bundled scripts
- Yes. scripts/extract.py and an extractor package run locally with Python 3, and tools/validate_skill.py checks the generated skill. Some formats need extra tools: poppler-utils, pypdf, pdfminer.six, docling, ebooklib, python-docx, Calibre for MOBI and AZW, and ocrmypdf for scanned PDFs.
- Network calls
- No calls are documented for the extractor, which the README says runs locally. The README says any text sent to a cloud-hosted model follows that provider's terms, and it mentions optional publishing of a generated skill to GitHub, private by default.
- API keys needed
- None named in the README.
These are facts read from the repo, not a security rating. Read the files before you install any skill.
What is inside Book to skill
| Skill or command | What it makes the agent do |
|---|---|
| book-to-skill | Extracts text from a book or folder of documents with a Python extractor, then has the agent write a skill from it with mental models, chapter files, a glossary, patterns and a cheatsheet. |
Run it as /book-to-skill ./my-book.pdf. The generated skill is written to ~/.agents/skills/<slug>/ and, under Claude Code, linked into ~/.claude/skills/<slug>/.
A good fit for
- Engineers who study technical books and want to ask the agent about them while they work
- Teams that want an author's frameworks applied during code review or design
- People who keep PDFs, EPUBs and notes and want a searchable skill from them
- Anyone who wants to load a book on demand instead of pasting it into the context
Skip it if
- Scanned books without text unless you first run OCR with ocrmypdf
- Redistributing a skill made from a copyrighted book, which the README advises against
- Setups where Python 3 and the per-format tools cannot be installed
How Book to skill works
The work is split in two. A deterministic Python extractor turns the source file into clean text and metadata. Then the agent follows the SKILL.md spec to write the skill: core mental models and a chapter index in SKILL.md, one file per chapter loaded only when needed, and glossary, patterns and cheatsheet files.
Afterwards you query the book by name, for example /my-book replication. The README advises keeping skills made from third-party books private and not redistributing them.
- Made by
- virgiliojr94
- Topic
- Research and data
- Works with
- Claude Code, GitHub Copilot, Amp, Hermes Agent, OpenCode, OpenClaw
How much context Book to skill uses
| The skill | 7,541 words in its SKILL.md file |
|---|---|
| Loaded at start | Only each skill's name and description. The full text loads when a task needs that skill. |
Word counts are measured from the repo every day. Reference files a skill loads on demand are not counted.
Is Book to skill still maintained?
| Last commit | , 3 days ago |
|---|---|
| Latest release | v1.4.0, 10 August 2026 |
| Stars gained in 7 days | Shown after 7 days of tracking |
| Stars gained in 30 days | Shown after 30 days of tracking |
| Open issues and pull requests | 39 |
| Forks | 3,609 |
Tools Book to skill works with
Similar skill packs
Planning with Files
A planning skill that keeps task_plan.md, findings.md and progress.md on disk so an agent resumes after /clear.
Scientific Agent Skills
177 research skills for databases, Python science packages, lab platforms and scientific writing.
HumanLayer skills
Six skills for CLAUDE.md, visual pull requests and agentic loops, from HumanLayer.
Questions about Book to skill
Is Book to skill free?
Yes. It is under the MIT license, which the README says covers the converter and not any book you process with it.
Which file types does Book to skill read?
PDF, EPUB, DOCX, Markdown, HTML, plain text and RTF, plus MOBI and AZW when Calibre is installed.
How do I use it?
Run /book-to-skill ./my-book.pdf, or give it a folder or glob and an optional skill name. Then query the new skill by its name.
Does it work with scanned PDFs?
Only after OCR. The README suggests running ocrmypdf on the file first.
Sources
Something wrong or out of date? Email [email protected] with a link to the source. Stats come from the public GitHub API and update every day.
For the maker
Made Book to skill? Claim this page to show readers you check it.
Add this badge to your README. The number updates every day.
[](https://scratchdb.com/skills/book-to-skill/)
Sharing on X? This page shows a share card with the pack's live numbers.
