Appearance
Books
FeynmanLM treats books differently from articles and papers. Instead of processing an entire book at once, it works chapter by chapter, letting you build your source library gradually as you read.
Where to Store Books
Save book PDFs to this iCloud Drive folder:
~/Library/Mobile Documents/com~apple~CloudDocs/Documents/FeynmanLM/Books/In Finder, this appears under iCloud Drive → Documents → FeynmanLM → Books. Create the folder if it doesn't exist.
Organize books into subfolders by title or author if you have many:
Books/
├── Deep Learning (Goodfellow)/
│ ├── deep-learning-ch01-intro.pdf
│ ├── deep-learning-ch02-linear-algebra.pdf
│ └── deep-learning-ch03-probability.pdf
├── The Pragmatic Programmer.pdf
└── Thinking Fast and Slow.pdfChapter Splitting
Long books need to be split into chapter-sized chunks. The app works best with PDFs in the 20–80 page range per source.
Why split by chapter:
- Your AI assistant gets more focused context when the source covers one topic
- You can process chapters as you read, rather than the whole book upfront
Splitting a PDF on Mac:
Open the book PDF in Preview, then:
- Go to View → Thumbnails to see the page sidebar
- Select the pages for a chapter (click + Shift-click to select a range)
- Drag the selected pages to the Finder folder, or use File → Export as PDF
Alternatively, use pdftk in Terminal:
bash
# Extract pages 45–78 (Chapter 3) from a book
pdftk book.pdf cat 45-78 output deep-learning-ch03-probability.pdfAdding a Book Chapter
- Save the chapter PDF to the Books folder in iCloud Drive.
- Open FeynmanLM. The chapter PDF appears in your Studio queue automatically.
- Select the chapter to load its content. It's now available to your AI assistant via MCP.
Tips
- Rename chapters descriptively:
deep-learning-ch03-probability-and-information-theory.pdfgives the AI better context thanchapter-03.pdf. - Skip appendices and references: Process only the chapters you're actively reading.
- Process gradually: Add chapters as you read them — no need to do the whole book upfront.