skills.rest
zachmayer
zachmayerCOMMUNITY
Explore all skills by this author →
Website𝕏@AI_Insight_SolsGitHub

photo-search

Search local Apple Photos libraries by GPS coordinates, text, and visual similarity.

2|Updated Oct 30, 2025
Search Tags:#indexing#search#macos#computer vision#photos#clip
One-click install
npx skills add https://github.com/zachmayer/skills --skill photo-search
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: photo-search
Source: https://github.com/zachmayer/skills/tree/main/.claude/skills/photo-search
Command: npx skills add https://github.com/zachmayer/skills --skill photo-search

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires click, numpy, open-clip-torch, osxphotos, pillow, pillow-heif, polars, scikit-learn, torch, and includes scripts (resource) components.

What problem does it solve?

This skill solves the difficulty of finding specific photos in large local Apple Photos libraries by enabling semantic, location-based, and visual similarity searches without needing to upload your private data to the cloud.

Core Features & Use Cases

  • Semantic Text Search: Find photos using natural language descriptions like "a framed picture on a wall" powered by CLIP embeddings.
  • Location-Based Discovery: Search for photos taken within a specific radius of a GPS coordinate, even if the photos lack explicit GPS tags by using visual neighbor voting.
  • Visual Similarity: Find photos that look like a reference image, useful for finding duplicates or related shots.
  • Use Case: Quickly locate all photos taken within 1km of a specific vacation spot or find all images that visually resemble a specific product or person.

Quick Start

Use the photo-search skill to index your library and then search for photos matching the description of a sunset over the ocean.

Frequently Asked Questions about photo-search

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I search my local Apple Photos library using natural language descriptions?▼

Semantic photo search uses CLIP embeddings to match natural language queries with image content. This allows you to find photos matching a sunset over the ocean by processing visual and text similarity directly on your macOS device.

Can I find photos in my macOS Photos library using GPS coordinates?▼

Yes, location-based photo discovery uses GPS coordinates to find images taken within a specific radius. It can even infer missing location data by using visual neighbor voting on surrounding photos from your library.

How do I find visually similar photos or duplicates in a local image library?▼

You can find visually similar photos or duplicates by using visual similarity matching. This compares a reference image against your local library to locate related shots or duplicate images without relying on cloud-based services.

Does semantic photo search on macOS require uploading my pictures to the cloud?▼

No, semantic photo search on macOS operates entirely locally using open-clip-torch and polars. It processes your library metadata and vector search operations on-device to ensure complete data privacy and high-performance retrieval.

What do I need to index and search Apple Photos with AI on macOS?▼

To index and search Apple Photos with AI, you need a macOS environment with osxphotos, open-clip-torch, and polars. These dependencies manage your library metadata and generate parquet-based indexes for high-performance vector search.

Related Skills

Explore More Data & Analytics Skills→
malue-aimalue-ai
36

apple-photos

Search and manage Apple Photos libraries on macOS using osascript.

Official
Intermediate
starlake-aistarlake-ai
209

parquet2csv

Convert Parquet files into CSV format for batch ETL pipelines.

Official
Intermediate
nadicodeainadicodeai

chroma

Index and search embeddings locally with a four-function vector store API.

Official
Intermediate
automatedigitalautomatedigital
1

clip

Classify images zero-shot and retrieve cross-modal matches via cosine similarity.

Official
Intermediate