Uber
Book Uber rides directly through your AI assistant.
MCP Uber Server
An MCP (Model Context Protocol) server for booking Uber rides through AI assistants.
Features
- OAuth 2.0 authentication with Uber
- Get price estimates for rides
- Request Uber rides
- Check ride status
- Cancel rides
Installation
Using npm (global installation)
npm install -g mcp-uber
Using npx (no installation required)
npx mcp-uber
Setup
Step 1: Create an Uber Developer Account
- Go to Uber Developer Dashboard
- Click "Sign in" and either:
- Use an existing Uber rider/driver account
- Create a new account specifically for development
💡 Tip: For organizations, create an email alias (e.g., [email protected]) instead of using a personal account for easier ownership transfer.
Step 2: Create a New App
- In the Developer Dashboard, click "Create App" (top right corner)
- Fill in the required information:
- App Name: Your application name
- Description: Brief description of what your app does
- Click "Create"
Step 3: Get Your API Credentials
- Navigate to your app in the dashboard
- Go to the Auth tab
- You'll find:
- Client ID: Public identifier for your app
- Client Secret: Private key (keep this secure!)
- Server Token: For server-side requests
Step 4: Configure OAuth Settings
- In the Auth tab, add your redirect URI:
- For local testing:
http://localhost:3000/callback - For production: Your actual callback URL
- For local testing:
- Select required scopes:
profile- User's basic profile informationrequest- Request rides on user's behalfride_request- View and manage active ride requests
⚠️ Note: The request scope is privileged and requires Uber approval for production use. During development, your account can use it without approval.
Step 5: Set Up Environment Variables
Create environment variables with your credentials (see Configuration section below)
Usage with Claude Desktop
Using npm (global installation)
Add to your Claude Desktop configuration (~/Library/Application Support/Claude/claude_desktop_config.json):
{
"mcpServers": {
"uber": {
"command": "mcp-uber",
"env": {
"UBER_CLIENT_ID": "your_client_id",
"UBER_CLIENT_SECRET": "your_client_secret",
"UBER_REDIRECT_URI": "http://localhost:3000/callback",
"UBER_ENVIRONMENT": "sandbox"
}
}
}
}
Using npx
Add to your Claude Desktop configuration:
{
"mcpServers": {
"uber": {
"command": "npx",
"args": ["mcp-uber"],
"env": {
"UBER_CLIENT_ID": "your_client_id",
"UBER_CLIENT_SECRET": "your_client_secret",
"UBER_REDIRECT_URI": "http://localhost:3000/callback",
"UBER_ENVIRONMENT": "sandbox"
}
}
}
}
Available Tools
- uber_get_auth_url - Get OAuth authorization URL
- uber_set_access_token - Set user's access token
- uber_get_price_estimates - Get price estimates for a ride
- uber_request_ride - Request an Uber ride
- uber_get_ride_status - Check ride request status
- uber_cancel_ride - Cancel a ride request
OAuth Flow
- Use
uber_get_auth_urlto get the authorization URL - User visits the URL and authorizes your app
- After callback, exchange the code for an access token
- Use
uber_set_access_tokento store the token - Now you can make API calls
Configuration
Environment Variables
The MCP server requires the following environment variables:
UBER_CLIENT_ID: Your Uber app client IDUBER_CLIENT_SECRET: Your Uber app client secretUBER_REDIRECT_URI: OAuth callback URL (default:http://localhost:3000/callback)UBER_ENVIRONMENT: Eithersandboxorproduction(default:sandbox)
Testing Your Integration
-
Use sandbox mode for testing:
- Set
UBER_ENVIRONMENT=sandboxin your environment - Sandbox mode simulates ride requests without real drivers
- Perfect for development and testing
- Set
-
Test the OAuth flow:
- Use the
uber_get_auth_urltool to get an authorization URL - Visit the URL and authorize your app
- After authorization, Uber will redirect to your callback URL with a code
- Exchange the code for an access token (you'll need to set up your own callback handler)
- Use
uber_set_access_tokento store the token in the MCP server
- Use the
-
Setting up a callback handler:
- For testing, you can use a simple Express server (see
examples/oauth-server.jsin the GitHub repo) - For production, implement a secure callback handler in your application
- The callback URL must match exactly what's configured in your Uber app
- For testing, you can use a simple Express server (see
Important Notes
Sandbox vs Production
-
Sandbox Mode (default):
- Simulated rides and drivers
- No real charges
- Perfect for testing
- Limited to your developer account
-
Production Mode:
- Real rides and charges
- Requires Uber approval for privileged scopes
- Must pass Uber's review process
Security Best Practices
- Never commit credentials: Keep your Client Secret secure
- Use environment variables: Don't hardcode credentials
- Implement proper token storage: The current in-memory storage is for demo only
- Validate redirect URIs: Ensure your callback URLs are properly configured
API Limitations
- Rate limits apply (check Uber's documentation)
- Privileged scopes require approval for production use
- Sandbox mode has some limitations compared to production
Troubleshooting
- "Invalid scope" error: Your app needs approval for privileged scopes in production
- "Invalid redirect URI": Make sure your redirect URI exactly matches what's configured in the Uber dashboard
- "Unauthorized" errors: Check that your access token is valid and not expired
Resources
관련 서버
mcp-atomictoolkit
An MCP-compatible server providing atomistic simulation capabilities through ASE, pymatgen, etc.
N.I.N.A. Advanced API
Control the N.I.N.A. (Nighttime Imaging 'N' Astronomy) software through its Advanced API.
HashLock OTC
OTC crypto trading with HTLC atomic settlement on Ethereum and Bitcoin — create trades, lock assets, and settle trustlessly via AI agents
Mercury MCP
Mercury MCP lets you ask questions about your banking balances, transactions, cards/recipients, and more
Ingero
eBPF-based GPU causal observability agent with MCP server. Traces CUDA Runtime/Driver APIs via uprobes and host kernel events via tracepoints to build causal chains explaining GPU latency. 7 MCP tools for AI-assisted GPU debugging and root cause analysis. <2% overhead, production-safe.
Jupiter Solana MCP Server
A comprehensive MCP (Model Context Protocol) server for interacting with Jupiter Protocol on Solana. Features token swaps, search, portfolio management, and intelligent error diagnostics.
MCP Cookie Server
Provides positive reinforcement to LLMs by awarding 'cookies' as treats.
FAIM Time-Series forecasting
Zero-shot Time-Series forecasting with foundation time-series models
Github MCP Server Java
A production-ready MCP server that connects any MCP-compatible AI agent to the GitHub API. Manage repositories, issues, pull requests, and search — all through natural language.
MCP Minecraft Remote
Remotely control a Minecraft Java Edition server using the Model Context Protocol (MCP).