arabiclisted

# /arabic — find & fetch the best edition of an Arabic Islamic book Backing tool location: `~/.claude/tools/arabic-library` (installed by this project's install script — see the repo README). Resolve `~` to the actual home directory for the current OS before running anything: - Windows PowerShell: `$env:USERPROFILE` - macOS/Linux: `$HOME` Always run commands with cwd set to that install directory. For the Python interpreter, try in order: `python`, then `python3`, then (Windows only) `py`. Use whichever one actually resolves on this machine. ## Scope (do not deviate from this) 1. **OpenITI first** — a CC0/public-domain corpus, safe to fetch freely. 2. **Archive.org second** — public API, only items already hosted as PDFs there. 3. **Local Shamela conversion — last resort only, and only from a file the user already has on disk.** Never fetch anything from shamela.ws or any Shamela server over the network. If the book isn't in OpenITI or Archive.org, ask the user whether they already have it downloaded in their official Shamela app, and if so, ask for the path to that `.db`/`.sqlite` file to convert. Do not attempt to scrape or download it from Shamela yourself. ## Procedure **1. Resolve what the user means.** Take whatever the user typed after `/arabic` (English transliteration, full Arabic title, author name, or a rough/partial name). Run: ``` python -m arabic_books search "<their text, as given>" ``` If that returns nothing useful, retry with varia