@RituWithAI: Someone just built a tool that turns any book into a skill your AI agent can use forever. One command. Any PDF, ePub, o…

Someone just built a tool that turns any book into a skill your AI agent can use forever. One command. Any PDF, ePub, or long document. A structured SKILL.md file that Claude Code, Cursor, or any MCP-compatible agent loads automatically whenever it's relevant. It's called book-to-skill. And the idea is almost embarrassingly obvious in retrospect. Here's the problem it solves. You have a book. "Clean Code." "Designing Data-Intensive Applications." "The Pragmatic Programmer." "Site Reliability Engineering." Thousands of pages of hard-won knowledge from the best engineers in the world. Your AI coding agent has never read any of them. Every session, you start from scratch. The agent doesn't know Martin Fowler's refactoring patterns. It doesn't know the SRE error budget philosophy. It doesn't know the Unix philosophy your codebase is built around. You prompt it every time. Or you don't, and it makes decisions without that context. book-to-skill reads the book for you and turns it into a skill your agent carries permanently. Here's how it actually works. Feed it any long-form document — a PDF, an ePub, a long markdown file, a scraped website. It chunks the content intelligently, identifies the core principles and patterns, and distills them into a structured SKILL.md file formatted specifically for how AI coding agents read and apply knowledge. Not a summary. Not a Q&A. A skill — organized into the sections that Claude Code, Cursor, and Codex are trained to read: when to use this skill, key principles, common patterns, anti-patterns, examples, and decision frameworks. Your agent now carries the essence of the book in a form it can actually apply while coding. Here's what this unlocks. "Clean Code" as a skill: your agent now flags functions longer than 20 lines, names variables with intention-revealing names, and keeps the boy scout rule in mind on every edit. "Designing Data-Intensive Applications" as a skill: your agent reasons about consistency models, replication strategies, and data pipeline design the way Kleppmann does. "The Pragmatic Programmer" as a skill: your agent applies DRY, YAGNI, and the broken window theory to every codebase it touches. One book. One command. Permanent knowledge upgrade for every agent session going forward. Three commands. Your agent just read Clean Code. 4 GitHub stars. Day one. This one is going to grow fast as the skills ecosystem discovers it. 100% Open Source. MIT License. GitHub link in the comments