Guide · Agent skill
How to use the Book Mirror agent skill (book-mirror)
Updated · written with AI from the project's own documentation and checked against our catalogue data
Book Mirror is an agent skill that reads a book you supply, such as an EPUB or PDF, and writes a personalized chapter-by-chapter analysis. For each chapter it summarizes the content in detail and then offers gentle observations linking the ideas to your own situations, people and patterns, drawn from your personal knowledge base. This guide is for readers who may never have used agent skills; by the end you will know what the skill does, how to install it, how to trigger it, and what it needs to run.
Book Mirrorgarrytan/gbrain · MIT Download the skill Quick start
- 1Download the ZIP. Download the book-mirror ZIP from this page.
- 2Install the skill. Unzip it into ~/.claude/skills/ or a project's .claude/skills/ folder, or upload it under Settings → Capabilities → Skills in the Claude apps.
- 3Set up the requirements. Make sure gbrain, Python 3 with beautifulsoup4 and lxml, unzip and pdftotext are available, and that your brain is populated.
- 4Provide your book. Supply the EPUB or PDF file yourself, since the skill does not download books.
- 5Ask in plain language. Use a request such as "mirror this book" or "apply this book to my life" along with the file.
- 6Review the output. Open media/books/<slug>-personalized.md, read the chapter analyses, and check it before sharing.
What is an agent skill and what does book-mirror do?
An agent skill is a folder with a SKILL.md file that tells an AI agent how to perform a specific task. When the skill is installed, the agent can follow its instructions when you ask for that kind of work.
Book Mirror takes a book and produces one brain page at media/books/<slug>-personalized.md. Each chapter has two parts: The Chapter, a detailed summary of the content, and The Mirror, observations that connect the ideas to your actual life, using context from your brain.
- The Chapter: a detailed account of what the chapter says.
- The Mirror: a friend-like observation of parallels to your own situations, people and patterns.
- The reader decides what to do about any observation. The skill is not meant to rearrange your life or assign homework.
- Optional: a PDF version of the page, rendered through the brain-pdf skill.
Install the book-mirror skill
Download the book-mirror ZIP from this page. You can install it in one of two ways, depending on where you use your agent.
For Claude Code, unzip the file into your user skills folder ~/.claude/skills/ for use everywhere, or into a project's .claude/skills/ folder for use in one project only. For the Claude apps, upload the ZIP under Settings → Capabilities → Skills.
- Claude Code, all projects: unzip into ~/.claude/skills/
- Claude Code, one project: unzip into the project's .claude/skills/ folder
- Claude apps: Settings → Capabilities → Skills, then upload the ZIP
unzip book-mirror.zip -d ~/.claude/skills/Check the requirements before you start
The skill depends on several tools that you need on your computer and in your brain. The source lists these requirements; install any that are missing before running the skill.
The skill is built around a personal knowledge base. Without a populated gbrain brain, the mirror has little to connect to.
- gbrain command-line tool with a populated personal brain
- Python 3 with beautifulsoup4 and lxml (for EPUB text extraction)
- unzip and pdftotext (poppler) command-line tools
- Optional: the brain-pdf skill, for PDF output
pip3 install beautifulsoup4 lxmlHow to trigger book-mirror with natural requests
You do not need a special command. Ask in plain language and include the book file. The skill responds to phrases such as the ones below, so you can use wording close to these.
Provide the path to your EPUB or PDF. The skill does not download books; you must supply the file yourself.
- "Make a personalized version of this book" (with the file)
- "Mirror this book"
- "Two-column book analysis"
- "Apply this book to my life"
- "How does this book apply to me?"
Extract the text from your book
Before analysis, the skill turns the book into one text file per chapter. For EPUB files it unzips the book and strips the HTML using Python. For PDF files it runs pdftotext.
After extraction, check the chapter files to separate real chapters from front matter such as the table of contents, copyright page and acknowledgments. The skill also suggests a quality check, including a word count for each chapter.
- EPUB: unzip the file, then strip HTML to text per chapter with BeautifulSoup
- PDF: run pdftotext with the -layout option, then split by chapter heading
- If a PDF has no embedded text, the source suggests OCR through brain-pdf or another vision tool
pdftotext -layout path/to/book.pdf full.txtWhat the output looks like and where it is saved
The result is a single brain page at media/books/<slug>-personalized.md, where the slug is a kebab-case name for the book. Each chapter is laid out as a top-aligned HTML table or as stacked sections. The source says plain markdown pipe tables are avoided because they misalign uneven columns.
If you want a PDF, the optional brain-pdf skill can render the page for delivery.
- Output path: media/books/<slug>-personalized.md
- Layout: top-aligned HTML table or stacked sections
- Optional PDF via brain-pdf
Safety notes: scripts, network and personal data
This skill runs scripts, so read what it does before you run it. According to the source it does not use the network. It does write files into your brain and reads your personal notes and people pages to build the mirror.
The analysis draws on personal notes and people pages from your brain. Review the output before you share it with anyone.
- Runs scripts: yes
- Uses network: no
- Writes to media/books/ in your brain
- Reads personal notes and people pages, so check the output before sharing
- Only use books you have the right to use; the skill does not download books
What people use it for
- Reading a self-help book and noting where its ideas match patterns already in your notes
- Working through a relationship book and seeing parallels with people pages in your brain
- Reviewing a business book against your own projects and decisions
- Keeping a personal, searchable record of how each chapter connected to your life
Frequently asked questions
How do I install the book-mirror skill?
Download the book-mirror ZIP from this page. Unzip it into ~/.claude/skills/ for all projects or into a project's .claude/skills/ folder for one project. In the Claude apps, upload the ZIP under Settings → Capabilities → Skills.
What do I need to run book-mirror?
You need the gbrain command-line tool with a populated personal brain, Python 3 with beautifulsoup4 and lxml, and the unzip and pdftotext (poppler) tools. The brain-pdf skill is optional and only needed for PDF output.
Is book-mirror free to use?
The source lists the skill's licence as MIT. The source does not mention pricing, so check the directory page for any other terms.
Does book-mirror download books for me?
No. The skill intentionally does not download books, so you must supply the EPUB or PDF file. Only use books you have the right to use.
Can I just get a plain summary of a book instead?
Book Mirror is built around personal mirroring, and the source says to use a different skill if you only want a plain summary. Book Mirror is not a summary tool or an advisor telling you what to change.
Why does the output need checking before I share it?
The analysis pulls personal notes and people pages from your brain. Review the output before sharing it, because it may contain personal context.
New agent skills, repos and AI tools — every week
The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.
Our verdict
Book Mirror suits readers who keep a personal knowledge base in gbrain and want self-help, relationship or business books connected to their own life, and who are comfortable running command-line tools. If you only want a plain book summary, or you do not want personal context analysed, pick a summary tool instead.
Download the skill