Provides tools for fetching restaurant reviews and ratings through the Google Places API, allowing users to compare TGO Yemek ratings with Google's data.
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Food402What pizza restaurants are near me?"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Food402 MCP Server
An MCP (Model Context Protocol) server that enables AI assistants to order food from TGO Yemek. Simply chat with your AI assistant to browse restaurants, build your order, and complete checkout. Works with Claude, ChatGPT (Developer Mode), and Codex CLI via MCP.
Clawdbot / OpenClaw (AgentSkills)
Food402 is also available as an AgentSkill on ClawHub for use with Clawdbot (OpenClaw), Claude Code, Cursor, Codex, Gemini CLI, and other AgentSkills-compatible tools.
Install via ClawHub:
Or manually: Copy the skill/ folder to ~/.openclaw/skills/food402/ and configure your credentials in ~/.openclaw/openclaw.json.
Local MCP Server (npm package)
Claude Desktop
Add to your Claude Desktop config (~/Library/Application Support/Claude/claude_desktop_config.json on macOS):
Replace credentials with your own. GOOGLE_PLACES_API_KEY is optional (see Optional Features).
Claude Code
For project-specific installation with Claude Code:
This automatically adds food402 to your .mcp.json. Open the file and update your credentials:
GOOGLE_PLACES_API_KEY is optional (see Optional Features).
Codex CLI (Terminal)
Codex reads MCP servers from your global config at ~/.codex/config.toml.
Option A: Via CLI
Option B: Manual config
Add to ~/.codex/config.toml:
Prerequisites
Account Setup Required
Before using this MCP server, you must have a TGO Yemek account with:
TGO Yemek account - Create one at tgoyemek.com if you don't have one
Payment card saved to your account - The checkout process requires a saved card; you cannot enter card details during ordering
At least one delivery address saved (recommended) - You can add addresses through the MCP, but having one pre-configured makes ordering faster
Quick Start: Ordering Flow
Here's the typical workflow when ordering food through the AI assistant:
1. Select Delivery Address
2. Find Restaurants
3. Browse Menu & Add Items
4. Review & Checkout
5. Place Order
Note: A browser window will open for 3D Secure verification. Complete the verification to finalize your order.
6. Track Order
Available Tools
Tool | Description | Parameters |
| Get user's saved delivery addresses | None |
| Select delivery address (must call before ordering) |
|
| Search restaurants near a location |
|
| Search restaurants and products by keyword |
|
| Get restaurant's full menu |
|
| Get product customization options |
|
| Get "goes well with" suggestions |
|
| Add items to cart |
|
| Get current cart contents | None |
| Remove item from cart |
|
| Clear entire cart | None |
| Get list of all cities for address selection | None |
| Get districts for a city |
|
| Get neighborhoods for a district |
|
| Add a new delivery address |
|
| Get user's saved payment cards (masked) | None |
| Get basket ready for checkout with payment context | None |
| Set order note and delivery preferences |
|
| Place order with 3D Secure (opens browser for verification) |
|
| Get user's order history with status |
|
| Get detailed order info including delivery status |
|
Optional Features
Google Reviews Integration
The get_google_reviews tool allows you to fetch Google Reviews for restaurants, enabling comparison between TGO ratings and Google ratings. This feature is optional and requires a Google Places API key.
To enable Google Reviews:
Get a Google Places API key from the Google Cloud Console
Enable BOTH APIs for your project (both are required):
Places API (New) - Used for text search to find restaurants
Places API - Used for fetching place details and reviews
Add the key to your MCP server configuration:
Without the API key: The tool will return a helpful message indicating the feature is unavailable but won't cause any errors. The rest of the food ordering functionality works normally without it.
Development
Repository Structure
Local Server Development
Testing
The project includes integration tests and security tests using Vitest.
Setup for tests:
Copy
test/.env.test.exampletotest/.env.testAdd your TGO credentials to
test/.env.testRun
npm test
Tests cover:
Address management (cities, districts, neighborhoods)
Restaurant discovery and menu browsing
Basket/cart operations
Payment card retrieval and checkout readiness
Order history
Security checks for credential handling
License
MIT