Read Calibre catalog data via calibredb over a Content server, and run one-book analysis workflow that writes HTML analysis block back to comments while caching analysis state in SQLite. Use for list/search/id lookups and AI reading pipeline for a selected book.