Smart Photo Journal MCP Server is an MCP server that connects Claude Desktop to the local Apple Photos library on macOS. Designed for Apple Photos users and creative personal journalers, it enables large language models to inspect and filter local photography collections without uploading raw assets to third-party cloud infrastructure. The server provides dedicated tools to query images based on geographical locations, system labels, keywords, and tagged people. It also supports fuzzy matching so users can locate people or places even with approximate name queries. Beyond catalog searching, the server includes metadata analysis capabilities that inspect photo timestamps and dates to uncover photography habits, such as identifying the days or times of week when the user takes the most pictures. All photo lookups occur on the local machine once system permissions are granted, letting users reminisce, prepare memory albums, or pull up specific snapshots directly through natural language conversations.
Category: Design, Media & Creative
Tags: journaling, macos, media-management, photos
Visit Smart Photo Journal MCP Server
uv. 2. Clone the repository: bash git clone https://github.com/Siddhant-K-code/memory-journal-mcp-server.git cd memory-journal-mcp-server 3. Install dependencies: bash uv sync 4. Add the server to your claude_desktop_config.json file: json { "mcpServers": { "smart-photo-journal": { "command": "/Users/<YOUR_DEVICE_USERNAME>/.local/bin/uv", "args": [ "--directory", "/Users/<PATH_TO_CLONED_DIR>/memory-journal-mcp-server", "run", "server.py" ] } } } 5. Replace <YOUR_DEVICE_USERNAME> and <PATH_TO_CLONED_DIR> with your actual paths. 6. Restart Claude Desktop and grant permission to access your Photos library when prompted.Part of MCP Servers
Smart Photo Journal MCP Server is an integration tool that exposes your local macOS Photos library to Model Context Protocol clients. It lets AI assistants search images by location, label, and person name, as well as analyze photography frequency and shooting patterns.
The server is built specifically for macOS because it connects directly to the native macOS Photos library. It requires a local macOS environment and the uv package manager to execute the Python server script.
The server is documented for Claude Desktop via its standard configuration file, but it works with any MCP client that supports standard local stdio server process execution.
No, the server executes locally on your machine. Photos and metadata are processed through local tools, and the operating system displays a standard prompt requesting permission to allow the local server to access your Photos library.