virgiliojr94/book-to-skill
A tool that turns technical book PDFs into on-demand reference skills for AI coding agents
book-to-skill converts technical book files in formats like PDF, EPUB, and DOCX into structured skills that AI coding agents such as Claude Code, GitHub Copilot CLI, and Amp can load on demand. Instead of dumping the whole book into an agent's context every time, it splits the content into core concepts, per-chapter files, a glossary, a pattern list, and a cheatsheet, loading only the relevant chapter when asked. On real books, this measured 24x to 51x fewer tokens than dumping the whole book into context to answer a single question.
- agent-skills
- ai-agents
- book-to-skill
- context-engineering
- document-processing
- edtech
- knowledge-base
- knowledge-management
- llm
- pdf-to-markdown
- rag
- self-study
What it does
- Point it at a file, folder, or glob of source documents, and a deterministic Python extractor pulls out clean text and metadata from formats like PDF, EPUB, DOCX, HTML, RTF, and MOBI.
- An AI agent then follows the instructions in a spec file called SKILL.md to organize that content into mental models, per-chapter files, a glossary, a patterns file, and a decision cheatsheet.
- Chapter files load only when a related question is asked, so the skill loaded in an agent's context stays small most of the time.
- The same approach works beyond books: internal documentation, brand guidelines, stacks of research papers, and specs or standards that people reference often but never memorize.
- It follows the open Agent Skills standard, so the same SKILL.md format works across GitHub Copilot CLI, Amp, and Claude Code.
Why it matters
People often read a technical book once and forget it months later, and asking an AI about it usually leads to made-up answers because the book was never actually loaded. This tool matters because it lets developers pull real, structured content from books or documents they own, right when they need it during actual work.
Terms in this repo
- Agent Skills · an open standard format that defines how AI agents load and use add-on capabilities
- SKILL.md · the spec file that instructs an agent how to build and use a skill
- token · the basic unit of text an AI model processes, tied directly to cost and speed
- OCR · technology that recognizes text in scanned page images and converts it into readable text
- on-demand loading · loading data only when it is actually needed, rather than all at once
Repository description (English)
Turn any technical book PDF into a Claude Code skill — ready to study, reference, and use while you work.
Open on GitHubCoverage
Trending repos
- openai/codexOpenAI's coding agent, now runnable straight from your terminal
- cordiverse/cordisA TypeScript framework for snapping feature pieces on and off whenever you need them
- ripienaar/free-for-devA single repo lists every free-tier cloud, API and collaboration tool developers actually need
- Wei-Shaw/sub2apiA relay server that lets one Claude/OpenAI/Gemini/Grok subscription be shared by many users
- multica-ai/andrej-karpathy-skillsA single instruction file that stops AI coding assistants from making silent bad guesses
- eneskirca/nodetermA workspace that puts scattered terminal tabs and AI coding agents on one draggable map
- affaan-m/ECCA toolkit that teaches AI coding agents to work like disciplined engineers
- n8n-io/n8nn8n lets you build automation workflows and AI agents by dragging nodes on a canvas, with code as a fallback
Latest from METAL LAB
- MCP's new roadmap enables stateless horizontal scaling for servers
- Turning Technical Book PDFs Into AI Skills Cuts Token Use Up to 51x
- Founders Are Pulling All-Nighters Just to Keep Up With Their AI Agents
- AI boss Luna made its first firing decision only after a human stepped in
- Anonymous model Ox Alpha matches GLM-5.2 on all 60 tokenizer tests