Learn VaultBook
inside and out

The 4-page carousel sits at the top of your workspace and rotates through: Suggestions (upcoming scheduled entries + weekday reading patterns from the last 4 weeks), Recently Read (up to 100 deduped entries with timestamps), Recent Files, and Recent Tools. It learns a personalized relevance distribution across your entire library — the more you use VaultBook, the smarter it gets.

Type any question in natural language. VaultBook searches across titles (8×), labels (6×), inline OCR text (5×), body content (4×), sections (3×), main attachments (2×), and section attachments (1×) — all weighted for relevance. Results are paginated 6 per page with navigable prev/next. The system automatically warms up attachment text for the top 12 candidates in the background.

Upvote or downvote any search result or related entry suggestion. Votes persist across sessions and directly influence sort ordering — upvoted items float to the top (+1M score boost), downvoted items sink. The add-to-note button lets you prefill a new entry from any result. An educational confirmation dialog (which you can disable) explains how voting improves your library over time.

When you're browsing any entry, Related Entries analyzes contextual similarity and surfaces notes you might want to revisit. Results appear with a fade-in animation and are paginated. Use Reddit-style upvote/downvote on each pair to train the relevance model — votes are saved to your repository and improve suggestions across sessions.

Every inline image in your entries is automatically OCR'd. Extracted text is cached per-entry and fully indexed, so you can search for words that only exist inside a photograph, screenshot, or scan. The top 12 QA results also trigger background OCR warm-up if the entry hasn't been processed yet — it all happens silently.

As you type in the main search bar, real-time dropdown suggestions appear — matching titles, details, labels, attachment names, and content. Your past queries are remembered so you can recall previous searches with a single tap.

A full contenteditable editor with: bold, italic, underline, strikethrough, ordered & unordered lists, headings H1–H6, font family selector, case transformation (UPPER, lower, Title, Sentence), text color & highlight pickers, tables with a visual size picker and right-click context menu for row/column operations, code blocks with language labels, callout blocks with accent bars, inline images, links, and built-in Markdown rendering via marked.js.

Break long notes into collapsible accordion sections. Each section has its own title, rich text body, and dedicated file attachments. Clip count badges show how many files each section holds. Sections are ideal for research papers, project phases, or any note that benefits from internal structure.

Every entry carries: title, body (rich text or plain), labels, page path (hierarchical placement), attachments (per-entry and per-section), sections, favorite toggle, protected/encrypted status, due date, expiry date, repeat/recurrence, and automatic created/updated timestamps. Every field is searchable and filterable.

Every edit creates a version snapshot stored in the /versions directory with 60-day retention. Browse your history in a modal sorted from newest to oldest, and restore any previous state with one click. Access history from the ⏳ button on any entry card.

While editing, VaultBook analyzes your content and suggests relevant labels as pastel-colored chips with counts. One click to apply — it speeds up tagging without interrupting your writing flow.

Create a page tree with unlimited parent/child nesting, disclosure arrows, drag-and-drop reordering, page icons, color dots, right-click context menus (rename, delete, move), and activity-based sorting. An "All Pages" root view lets you see everything at once.

Labels are color-coded pill tags in the sidebar — click to filter. Inline #hashtags in note content auto-generate Kanban columns in the board tool. Favorites (star toggle) get a dedicated quick-access panel in the sidebar with a compact scrollable list.

Open multiple independent entry list tabs at once. Each tab maintains its own sort, filter, pagination, and view state. The tab strip includes a + button to add new views — perfect for comparing research across different pages or labels.

Filter by file type (match any or match all), date field, and date range (any, 7 days, 30 days). Combine filters freely. Sort by multiple fields with a dropdown and toggle ascending/descending.

Each protected entry is encrypted independently with AES-256-GCM. Key derivation uses PBKDF2 with 100,000 iterations of SHA-256. Every encryption generates a unique random 16-byte salt and 12-byte IV. Each entry has its own password — compromising one never exposes the rest. Decrypted plaintext is held in memory only and never written to disk.

Activate the lock and the entire app blurs behind a full-page overlay. All pointer events and text selection are blocked — nothing is visible or interactable until you authenticate. Walk-away protection for shared environments.

VaultBook has no backend, no API calls, no telemetry, no analytics pings. It runs entirely in your browser from a local HTML file. No data ever travels over a network connection. Your documents stay on your device, period.

Auto-generates drag-and-drop columns from your labels and inline #hashtags. Cards stay synced with your notes — move a card and the entry updates, edit a note and the board reflects it.

Drop a CSV or TXT file and instantly see your data analyzed and visualized. Ideal for quick data exploration without leaving your workspace.

Follow feeds organized in folders. Stay current with blogs, news, and publications — all inside VaultBook without opening another app.

A centered overlay for rapid-fire, chat-style note capture. Great for brainstorms, meeting notes, and stream-of-consciousness writing.

Paste any web page URL and VaultBook creates a searchable note from it. Capture articles, references, and bookmarks directly into your library.

Trim silence, cut clips at precise points, and join multiple audio segments — all processed locally in your browser.

Combine multiple PDFs into one, split a PDF into individual pages, or compress scanned documents to reduce file size — no upload to any server.

Browse all attachments by type, entry, or page in the File Explorer. The Photo & Video Explorer lets you scan folders of media for a visual gallery experience.

Folder Analyzer visualizes disk space and file sizes at a glance. Password Generator creates strong passwords and copies with one click. Import from Obsidian lets you drop .md files to migrate notes instantly.

Attach files to entries and to individual sections within entries. All files are stored via the File System Access API in the /attachments directory with a JSON manifest (index.txt). Right-click context menus, pending/failed state indicators, and a reindex button (♻️) to rebuild the index at any time.

Pro unlocks deep text extraction from: XLSX/XLSM (SheetJS), PPTX slides (JSZip), PDF text layers (pdf.js), ZIP archive contents, and MSG Outlook emails (subject, from, body + nested attachments). All extracted text is indexed for search. Background warm-up ensures top results have their attachment text ready before you even scroll.

Pro goes beyond inline OCR — it extracts and OCR's images inside other file formats: images within ZIP archives, rendered PDF pages (scanned document support), embedded images inside DOCX (word/media/*), and embedded images inside XLSX (xl/media/*). If there's text in an image buried inside a document, VaultBook will find it.

A full calendar with tabbed day and week views over a scrollable 24-hour timeline. Task scheduling and prompts integrate with AI suggestions (upcoming entries in the next 48 hours). All data is persisted to disk so your schedule survives refreshes and sessions.

The sidebar always shows three time-oriented tabs: Recent (recently modified entries), Due (upcoming due dates), and Expiring (approaching expiry). A fourth Tools tab gives quick access to all built-in tools.

Three Canvas-rendered charts: label utilization pie chart, last 14 days activity line chart, and pages utilization pie. Plus attachment type breakdown chips and inline metric pills for entry counts, file counts, and total storage size — all computed locally.

The 🎲 sidebar widget surfaces a random note every hour with a "new pick" indicator — a gentle way to rediscover forgotten ideas. Companion widgets show a timetable ticker (upcoming events) and attachment ticker (recent file activity).