Émile
Personal wine cellar — list, add, recommend, scan and search 100k+ wines. Available via npx and as an OAuth-protected remote MCP at mcp.emile.wine.
Use Émile from your AI assistant
Émile ships an official MCP server. Ask Claude, ChatGPT, Cursor or any MCP-compatible client to add, scan or recommend bottles straight from your cellar. Two connection modes: remote OAuth (simplest, recommended) or local npx (for clients without OAuth support).
Prerequisites: an Émile account (sign up via the iOS app or at app.emile.wine).
What you can do
| Tool | Description |
|---|---|
| search_wines | Search the 100k+ catalog by name, producer, appellation |
| list_cellars | List your cellars |
| list_bottles | List bottles in a cellar |
| add_bottle | Add a bottle |
| add_bottles_batch | Add up to 100 bottles at once |
| update_bottle | Change status, notes, location, price, vintage, or move between cellars |
| delete_bottle | Remove a bottle |
| recommend_bottles | Pairing recommendation for a meal or based on drinking maturity |
| scan_label | Identify a wine from a label photo |
| scan_labels_batch | Up to 10 label photos at once |
Option 1 — Remote connection (OAuth)
The simplest path: Claude Desktop, ChatGPT and other recent clients support adding a remote MCP server via OAuth — no local config file, no API key to manage.
Claude Desktop
- Settings → Connectors → Add custom connector
- URL:
https://mcp.emile.wine/mcp - Click Connect → you'll be redirected to
app.emile.wineto sign in and authorize - Back in Claude, the connector is active. Try: "List my cellars"
ChatGPT (Developer Mode)
- Settings → Connectors → enable Developer Mode
- Create app → URL:
https://mcp.emile.wine/mcp - Authentication: OAuth (auto-detected)
- Complete the authorization flow (Émile sign-in) then test
Option 2 — Local install (npx)
For clients that don't yet support remote OAuth (Cursor, Continue, Cline…), the server runs locally via npm and talks directly to api.emile.wine using a personal API key.
1. Generate an API key from the app
Open Émile on your iPhone → Profile → Settings → API Keys → + Generate. Copy the emile_live_... key (shown only once).
2. Add the server to your config
Example for Claude Desktop — edit ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows):
{
"mcpServers": {
"emile": {
"command": "npx",
"args": ["-y", "@emile-wine/mcp-server"],
"env": {
"EMILE_API_KEY": "emile_live_xxxxxxxxxxxxxxxx"
}
}
}
}
For Cursor / Continue / others: same command npx -y @emile-wine/mcp-server + the EMILE_API_KEY env variable, adapt the syntax to your client's config format.
3. Restart your client
The MCP indicator should appear with emile listed in the connected servers.
Prompt examples
- "List the bottles in my main cellar"
- "Add a Château Pichon Lalande 2015, bought €95, slot A3"
- "What should we drink tonight with duck breast?"
- "Move the Chinon bottle from my garage cellar to the kitchen cellar"
- "Mark the last bottle of Cahors as consumed"
- "Scan this label: [photo]"
Privacy
Remote OAuth mode: no credential storage on the client side, authorization is revocable from your Émile account at any time. The server runs on our infrastructure in France behind a strict CSP, and every request is validated through a short-lived token.
Local npx mode: your API key stays on your machine. The server runs locally and talks directly to api.emile.wine. No third party in between — it's exactly like using the app, but via text.
In both cases, see the full privacy policy.
Also available on
- 📱 iOS App Store — the native Émile app
- 🌐 app.emile.wine — web signup + API key management
- 📦 npm @emile-wine/mcp-server
- 📋 Official MCP Registry — listed as
io.github.SebastienGal/emile-wine
Sources & support
- MCP spec: modelcontextprotocol.io
- Questions, feedback, bugs: [email protected]
संबंधित सर्वर
OraClaw Decision Intelligence
12 MCP tools with 19 ML algorithms for AI agents — bandits, solvers, forecasters, risk models. All under 25ms, deterministic.
Flightradar24
Track flights in real-time using Flightradar24 data.
Revelata deepKPI
Deep fundamental data from SEC filings, including operational KPIs not found on Bloomberg, built for your financial AI agents.
Bonnard
Ultra-fast to deploy agentic-first mcp-ready semantic layer. Let your data be like water.
SketchUp MCP Server
Control SketchUp with AI. MCP (Model Context Protocol) server that allows AI assistants like Claude, Cursor, and Gemini to programmatically create 3D models in SketchUp.
diagrams-mcp-server
MCP server for generating cloud architecture diagrams, flowcharts, sequence diagrams, and more — powered by mingrammer/diagrams, Mermaid, and PlantUML.
Uniswap Trader MCP
Automate token swaps on the Uniswap DEX across multiple blockchains.
Unphurl
URL intelligence for AI agents. 13 tools for security signals and data quality checks. Analyses URLs across 7 dimensions: redirect behaviour, brand impersonation, domain age, SSL/TLS, parked detection, URL structure, DNS enrichment. Risk score 0-100 with 23 configurable weights.
Image
Fetch and process images from URLs, local file paths, and numpy arrays, returning them as base64-encoded strings.
MCP Media Player
Control a media player via Home Assistant.