local-image-search_skill

Overview

This skill provides fast local image search on macOS using Spotlight (mdfind) or fd for filename-based lookups. It lets you locate images by name, date, GPS location, dimensions, camera metadata, or visual similarity and provides scripts to list, thumbnail, and copy results. Designed for quick, scriptable searches without manual browsing.

How this skill works

The skill wraps Spotlight queries (kMDItem keys) and optional fd commands to scan directories for image files and metadata. Scripts accept patterns and date ranges, filter by content type, GPS or pixel dimensions, and can pipe results into thumbnail generation, copying, or perceptual-hash similarity checks. Optional tools like exiftool improve metadata extraction and fd speeds up filename searches.

When to use it

• You need to find images quickly across a large local photo folder.
• You want to filter photos by creation or modification date (today, last week, month).
• You need to find images taken at or near specific GPS coordinates.
• You want to locate photos by camera make/model, resolution, or file type.
• You need a scriptable workflow to copy or generate thumbnails for a set of found images.

Best practices

• Index the directories you search with Spotlight and keep Spotlight enabled for reliable mdfind results.
• Install fd and exiftool for faster filename searches and richer metadata queries.
• Use targeted -onlyin paths to narrow searches and avoid scanning the whole disk.
• Combine date and content-type filters to reduce false positives (e.g., kMDItemContentTypeTree == 'public.image').
• Pipe search results into copy or thumbnail scripts rather than editing files in place to preserve originals.

Example use cases

• Find all iPhone photos (IMG_ prefix) in ~/Pictures and copy today’s results to a folder on the Desktop.
• List every image in a project directory and generate thumbnails for quick review.
• Search photos taken within latitude/longitude bounds to assemble images from a specific trip.
• Find images modified in the last seven days across multiple folders for recent edits.
• Detect visually similar images using perceptual hashes to remove duplicates or near-duplicates.