woof
Enables AI assistants to browse and search local photo galleries without uploading photos, using metadata-only indexing and local thumbnail serving.
README
<p align="center"><img src="assets/woof_large_850.png" alt="Woof" height="360"></p>
Woof — See your Photo Gallery in your AI assistant
Early preview release. Woof is functional but rough around the edges. Expect missing features, occasional errors, and breaking changes between releases. See the status section below.
Woof is the MCP App to "Où est Charlie ?", a media management system that keeps your photos (later movies and other media) exactly where they are — on your own drives — while giving you a beautiful, searchable gallery powered by your AI assistant (Claude, Goose...).
No cloud subscription. No proprietary lock-in. Your library, your way.
Où est Charlie?? That's "Where is Wally?" in French.
What makes it different
Most photo managers lock your library into a cloud service (Google Photos, iCloud) or require a database server that becomes a single point of failure. Woof takes a different approach:
- Conversation as your gallery. Woof connects to your AI assistant (Claude Desktop, ChatGPT, Goose…) and turns it into a full photo browser. Ask in plain language, get results inline. No separate app to learn.
- Privacy by design. Only metadata travels to your AI assistant — your actual photos are served locally by Woof. Your pictures are never uploaded to any AI service unless you explicitly ask.
- No database. Metadata lives as XMP sidecar files right next to your photos, plus lightweight JSON manifests. Move a drive, copy a folder — your entire organization travels with your photos.
- Open formats, forever. XMP is an ISO standard. JSON is universal. AVIF is royalty-free. Every tool you already use — Lightroom, darktable, ExifTool — can read your metadata today and long after OuEstCharlie is gone.
- Your photos are never touched. Woof reads your library as-is. It never modifies, moves, or deletes your original files. It also honors existing XMP metadata from Lightroom, darktable, or any other tool — rather than overwriting it.
- Works with your existing folder structure. Just point Woof at your photos folder. No migration, no reorganization required.
More about OuEstCharlie and Woof on the OuEstCharlie Blog
Installation
Woof runs as a local MCP server. It connects to your AI desktop client and exposes your photo library as a set of tools.
Option A — Bundle install (recommended but Claude Desktop only)
Connect to Claude Desktop
Download the latest ouestcharlie-woof.mcpb from the Releases page and double-click it. Claude Desktop will prompt you to install Woof in one click — no configuration file to edit.
See also in Woof Blog:
Option B — Manual uvx configuration
Prerequisites
Python packages of OuEstCharlie Woof are managed by uv and the command uvx. uv might be already available on your system.
Connect to Claude Desktop
Reference: Getting Started with Local MCP Servers on Claude Desktop
Open (or create) ~/Library/Application Support/Claude/claude_desktop_config.json and add or update mcpServers:
{
"mcpServers": {
"woof": {
"command": "uvx",
"args": ["--python", "3.12", "--from", "ouestcharlie-woof", "woof"]
}
}
}
Restart Claude Desktop. Woof will appear as an MCP integration, and the gallery will render as an interactive panel inside your conversation.
Connect to ChatGPT Desktop
NOTE: As o May 2026, ChatGPT is no longer supporting local MCP servers. Following is not longer available!
Connect to Goose
Reference: Goose MCP extensions documentation
Goose supports MCP servers via its extension system.
Either add through the user interface as a Custom Extension: <p align="center"><img src="assets/goose_custom_extension_setup.jpg" alt="Setup Woof extension in Goose" height="360"></p> <p align="center"><i>Setup Woof extension in Goose</i></p>
Or add the following to your Goose configuration (~/.config/goose/config.yaml):
extensions:
woof:
type: stdio
cmd: uvx
args: ["--python", "3.12", "--from", "ouestcharlie-woof", "woof"]
enabled: true
Other supported AI Assistants
Other clients support MCP Apps, for example VSCode Github Copilot or Codex.
See the MCP Extension Support Matrix
First Steps
1. Register your photos folder
Once Woof is connected to your AI client, ask it to register your photo folder:
"Add a local library to Woof pointing to /Users/yourname/Pictures"
Woof supports any folder on a local drive — including folders synced from iCloud Drive, OneDrive, or Google Drive, as long as the files are locally available.
2. Index your library
Trigger the indexer to scan your photos and build the metadata index:
"Index my local library"
Woof will launch the indexing agent, which will:
- Read EXIF/XMP metadata from each photo
- Write XMP sidecar files alongside your originals (never modifying the originals)
- Generate thumbnails and previews
- Build a fast index for querying
Indexing speed is roughly 10 to 100 seconds per 1,000 photos depending on format and hardware.
3. Start browsing
Once indexing is complete, just ask:
"Show me photos in Woof from last July"
"In Woof, show me pictures taken near Paris"
"Search Woof for photos with 'Tour Eiffel' in the description"
"How many photos do I have in Woof?"
The gallery panel will appear inline in your conversation with matching results.
<p align="center"><img src="assets/screenshot_2024-07-05.jpg" alt="Woof in Claude Desktop" height="500"></p> <p align="center"><i>Ouestcharlie Woof photo gallery inside Claude Desktop</i></p>
Storage
V1 supports local filesystem and cloud_mount libraries on macOS, Linux, and Windows:
- filsystem for a standard local hard drive or SSD, including local network drive (e.g. NAS)
- clound_mount for a folder synced from iCloud Drive, OneDrive, Google Drive, or Infomaniak kDrive — as long as files are downloaded and locally accessible
Native cloud storage (S3, Azure, GCS, OneDrive API) is planned for V2.
Status
Woof is an early preview targeting a focused V1 scope:
| Feature | Status |
|---|---|
| Local filesystem indexing (macOS, Linux, Windows) | Working |
| Mounted cloud drives (iCloud Drive, OneDrive, kDrive) | Working — files must be locally synced |
| JPEG, PNG, TIFF, HEIC, RAW support | Working (HEIC and RAW dependant on the build options) |
| Date, GPS bounding box, camera make and model, dimensions, tags search | Working |
| Full-text search on photo description | Working |
| Gallery view (Claude Desktop) | Working |
| Video support | Planned for V2 |
| Albums and smart filters | Planned for V2 |
| Share pictures with host (Claude Desktop, ChatGPT, Goose...) | Planned for V2 |
| Enrichment agents (faces, scene recognition) | Planned for V2 |
| Change detection / automatic re-indexing | Partial (added and removed pictures) |
| Mobile companion app | Planned for V3 |
| Native cloud libraries (S3, OneDrive, GCS…) | Planned for V3 |
What this means for you: V1 works well for browsing and searching a local photo library. If you hit a bug or unexpected behavior, please open an issue.
Privacy Policy
Woof is designed with privacy as a core principle.
- Data collected: Only photo metadata (EXIF, GPS coordinates, camera make/model, dates, file paths) is read and indexed. No account or personal information is collected.
- Data storage: All metadata is stored locally on your own device as XMP sidecar files and JSON manifests alongside your photos. No data is stored on any remote server.
- AI assistant: Only metadata and thumbnail images are sent to your AI assistant (Claude, ChatGPT, Goose…) when you perform a search. Your original photo files are never uploaded to any AI service unless you explicitly share them.
- Third parties: No metadata or usage data is shared with any third party.
- Retention: All data remains under your full control. Deleting the XMP sidecars and
.ouestcharlie/folders from your photo library completely removes all Woof metadata.
For privacy questions, please open an issue.
Support
Bug reports and feature requests: GitHub Issues
Developers' corner
For developer and architecture documentation, see README_DEV.md.
License
MIT license
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.