UPS MCP Server
Enables AI agents to integrate with UPS shipping and logistics capabilities, including package tracking with delivery status and transit information, and address validation for U.S. and Puerto Rico locations.
README
UPS MCP Server
A Model Context Protocol (MCP) server for UPS shipping and logistics capabilities. This server enables AI systems to seamlessly integrate with UPS API tools.
Users can integrate with the MCP server to allow AI agents to facilitate tracking events on their behalf, including tracking the status of a shipment, the latest transit screen, and expected delivery date and time. Agents will be authenticated using OAuth client credentials provided by the user after application creation on the UPS Developer Portal.
Usage
Prerequisites
- Obtain a Client ID and Client Secret: Create an application on the UPS Developer Portal to obtain your OAuth credentials – Client ID and Client Secret. (https://developer.ups.com/get-started?loc=en_US)
- Python 3.12 or higher
- Install uv (Python Package)
Environment Variables
CLIENT_ID- UPS Client IDCLIENT_SECRET- UPS Client SecretENVIRONMENT- Whether to point to Test (CIE) or Production (Accepted values: test, production)
Note: Your API credentials are sensitive. Do not commit them to version control. We recommend managing secrets securely using GitHub Secrets, a vault, or a password manager.
Execution
You can run the package using uvx:
uvx --from git+https://github.com/UPS-API/ups-mcp ups-mcp
To use an older version, you can specify the version number like so:
uvx --from git+https://github.com/UPS-API/ups-mcp@v1.0.0 ups-mcp
Popular Integrations
Here are sample config files for popular integrations. Different MCP Clients may require modification.
Claude Desktop
{
"mcpServers": {
"ups-mcp": {
"type": "stdio",
"command": "uvx",
"args": ["--from", "git+https://github.com/UPS-API/ups-mcp", "ups-mcp"],
"env": {
"CLIENT_ID": "**********",
"CLIENT_SECRET": "**********",
"ENVIRONMENT": "test"
}
}
}
}
GitHub Copilot in VS Code
{
"servers": {
"ups-mcp": {
"type": "stdio",
"command": "uvx",
"args": ["--from", "git+https://github.com/UPS-API/ups-mcp", "ups-mcp"],
"env": {
"CLIENT_ID": "**********",
"CLIENT_SECRET": "**********",
"ENVIRONMENT": "test"
}
}
}
}
Available Tools
-
track_package: Track a package using the UPS Tracking APIArgs:
- inquiryNumber (str): the unique package identifier. Each inquiry number must be between 7 and 34 characters in length. Required.
- locale (str): Language and country code of the user, separated by an underscore. Default value is 'en_US'. Not required.
- returnSignature (bool): a boolean to indicate whether a signature is required, default is false. Not required.
- returnMilestones (bool): a boolean to indicate whether detailed information on a package's movements is required, default is false. Not required
- returnPOD (bool): a boolean to indicate whether a proof of delivery is required, default is false. Not required
Returns:
- str: The response from the tracking capability, this is a string of json tracking data.
-
validate_address: Validate an address using the UPS Address Validation API for the U.S. or Puerto RicoArgs:
- addressLine1 (str): The primary address details including the house or building number and the street name, e.g. 123 Main St. Required.
- addressLine2 (str): Additional information like apartment or suite numbers, e.g. Apt 4B. Not required.
- politicalDivision1 (str): The two-letter state or province code, e.g. GA for Georgia. Required.
- politicalDivision2 (str): The city or town name, e.g. Springfield. Required.
- zipPrimary (str): The postal code. Required.
- zipExtended (str): 4 digit Postal Code extension. For US use only. Not required.
- urbanization (str): Puerto Rico Political Division 3. Only valid for Puerto Rico. Not required.
- countryCode (str): The country code, e.g. US. Required.
Returns:
- str: The json response body from the address validation API
UPS MCP server is still in active development. More tools coming soon!
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。