Skip to main contentSearch finds files by meaning, context, and intent rather than filenames or folders.
Global Search Interface
The global search modal provides quick access to your files and AI assistant. Open it from anywhere in Zark to search for files or ask questions.
When you first open the search modal, it displays your recent files organized by time—showing files from TODAY and THIS WEEK for quick access to your most recent work.
As you type your query, Zark AI provides suggestions to ask questions directly. Use ⌘ + ENTER to ask ZarkAI about your query, or continue typing to search for files.
Search results show matching files and folders in the MATCHES section, with timestamps indicating when files were last accessed. Results are ranked by relevance, not just filename matches.
Keyboard Shortcuts
- ↑↓ Navigate - Use arrow keys to move through results
- ⏎ Open file - Press Enter to open the selected file
- / Commands - Type ”/” to access command palette
- ⌘ + ENTER - Ask ZarkAI about your query
Command Palette
Type ”/” in the search interface to access the command palette, which provides quick access to Zark’s AI-powered features.
The command palette displays available commands:
/ask - Ask ZarkAI questions/image - Generate an image/crypto - Analyze on-chain data
Select a command to activate it. For example, selecting /ask prepares the interface for asking ZarkAI questions.
After selecting /ask, type your question and use ⌘ + ENTER to submit it to ZarkAI. The command palette also provides shortcuts for navigating, mentioning files, and opening files.
You can ask complex, multi-part questions using the /ask command. ZarkAI provides suggestions and can help refine your queries. Use ⌘ Navigate to move through suggestions, @ Mention files to reference specific files, or ⌘ + ENTER to submit your query.
How Search Works
Search combines multiple signals to find files:
- AI-generated tags
- File content analysis
- Semantic understanding
- Metadata and context
This allows finding files even when exact terms are unknown.
Search Components
Semantic Search
Semantic understanding interprets meaning, not just keywords. Queries are matched to content by concept and intent.
See How Indexing Works for how semantic indexing enables this.
Tag-Based Search
Tags generated during processing enable fast filtering and discovery. Multiple tags can be combined automatically from a single query.
See AI Auto-Tagging for how tags are created.
Content Analysis
Full content is analyzed, not just metadata. This enables finding information within files, not just about files.
Discovery Features
Search can surface:
- Similar files based on content
- Related topics across files
- Content with shared context
Search at Scale
Search works across large datasets:
- Thousands or millions of files
- No manual organization required
- Results ranked by relevance
