Search
Jump to a file by name or content from the sidebar search box.
AI
Preview — early access.
Overview
The search box at the top of the file tree finds a file two ways at once: by filename, from a persistent index that answers instantly, and by content, from a bounded scan across the workspace.
How it works
Type at least two characters and press Enter (or Escape to clear). Filename matches come back first since they're served from an index; content matches follow from a scan and can include a line number and a text preview. Results merge into one list, filename hits deduped against any content hit on the same path, with a note if the content scan was truncated for size.
Steps
- •
Click into the search box
It's above the file tree, in the sidebar.
- •
Type your query
At least two characters — file names and file content are both searched.
- •
Press Enter
Results list below, filename matches first.
- •
Open a result
Click any match to open that file — content matches jump you to the matching context.
- •
Clear when you're done
Press Escape or click × to reset the search.
Capabilities
- Instant filename search from a persistent index.
- Content search across workspace files, with line numbers and a preview.
- Merged, deduped results — a file matching both ways only appears once.
- Truncation notice when a content scan hits its size bound.
Limits & honest scope
- Content search is a bounded scan, not a full-text index — very large workspaces may see truncated results.
- Binary and very large files are excluded from the content scan.
- Results are workspace-scoped — search doesn't cross into a different workspace root.