
该项目已正式收录于 https://github.com/modelcontextprotocol/servers ,展示了社区构建和验证的 MCP 服务器。
概览
模型上下文协议(MCP)服务器,通过 OAuth 2.0 通过 PKCE 认证访问 NetSuite 数据。可无缝兼容任何兼容 MCP 的客户端,包括 Claude Code、Cursor IDE 和 Gemini CLI。 NetSuite 提供了官方的 AI 连接器套件应用,支持基于 AI 驱动的 NetSuite 数据交互。然而,NetSuite 的 AI 连接器目前仅支持: - 通过Anthropic的网页界面获得Claude- 通过自定义GPT连接的ChatGPT问题:使用兼容MCP工具如Claude Code、Cursor IDE、Windsurf或其他CLI/IDE环境的开发者无法利用NetSuite的AI能力,因为没有MCP服务器实现。 - 提供MCP客户端(Claude代码、光标、Gemini CLI等)之间缺失的桥梁。
概述(AI 翻译)
模型上下文协议(MCP)服务器,通过 OAuth 2.0 通过 PKCE 认证访问 NetSuite 数据。可无缝兼容任何兼容 MCP 的客户端,包括 Claude Code、Cursor IDE 和 Gemini CLI。
动机与背景
NetSuite 提供了官方的 AI 连接器套件应用,支持基于 AI 驱动的 NetSuite 数据交互。然而,NetSuite 的 AI 连接器目前仅支持:
- 通过Anthropic的网页界面访问Claude
- 通过自定义GPT连接的ChatGPT 问题在于:使用兼容MCP工具的开发者,如Claude Code、Cursor IDE、Windsurf或其他CLI/IDE环境,无法利用NetSuite的AI能力 因为没有MCP服务器的实现。
该MCP服务器通过以下方式解决了这一空白:
- 提供MCP客户端(Claude Code、Cursor、Gemini CLI等)与NetSuite的AI连接器之间缺失的桥梁
- 启用与 NetSuite AI 连接器完全相同的功能,但可通过任何兼容 MCP 的客户端访问
- 允许开发者直接在开发环境中使用自然语言与NetSuite数据交互
- 保持相同的安全标准(OAuth 2.
本页由 AI 翻译,技术细节请以原文为准。
README
以下为原文 README 节选,完整内容请查看原文。
NetSuite MCP Server
A Model Context Protocol (MCP) server providing access to NetSuite data through OAuth 2.0 with PKCE authentication. Works seamlessly with any MCP-compatible client including Claude Code, Cursor IDE, and Gemini CLI.
Motivation and Context
NetSuite provides an official AI Connector SuiteApp that enables AI-powered interactions with NetSuite data. However, NetSuite’s AI Connector currently only supports:
- Claude via Anthropic’s web interface
- ChatGPT via custom GPT connections The problem: Developers using MCP-compatible tools like Claude Code, Cursor IDE, Windsurf, or other CLI/IDE environments cannot leverage NetSuite’s AI capabilities because there’s no MCP server implementation.
This MCP server solves that gap by:
- Providing the missing bridge between MCP clients (Claude Code, Cursor, Gemini CLI, etc.) and NetSuite’s AI Connector
- Enabling the exact same functionality that NetSuite’s AI Connector provides, but accessible through any MCP-compatible client
- Allowing developers to interact with NetSuite data using natural language directly within their development environment
- Maintaining the same security standards (OAuth 2.0 with PKCE) required by NetSuite’s official AI Connector
In essence, this MCP server brings NetSuite’s AI capabilities to the broader MCP ecosystem, allowing developers to query business data, generate reports, and automate NetSuite operations without leaving their IDE or CLI.
Features
- ✅ OAuth 2.0 with PKCE - Secure authentication without client secrets
- ✅ Automatic Token Refresh - Tokens refresh automatically before expiration
- ✅ Environment Variable Support - Configure credentials once in your MCP config
- ✅ Session Persistence - Authentication survives server restarts
- ✅ Universal MCP Integration - Works with Claude Code, Cursor IDE, Gemini CLI, and other MCP clients
- ✅ NetSuite MCP Tools - Access to all NetSuite MCP capabilities (SuiteQL, Reports, Saved Searches, etc.)
- ✅ Modular Architecture - Clean, maintainable codebase following single-responsibility principle
Quick Start
1. NetSuite Setup
Step 1: Install NetSuite AI Connector SuiteApp
Before creating the integration record, you must install and configure the NetSuite AI Connector SuiteApp:
Important: The NetSuite AI Connector SuiteApp is required for MCP functionality. Without it, the MCP tools will not be available even after authentication.
Step 2: Create OAuth Integration Record
After installing the SuiteApp, create an integration record:
- Navigate to Setup > Integration > Manage Integrations > New
- Fill in the details:
- Name: “MCP Server Integration”
- OAuth 2.0: Checked Authorization Code Grant Checked Public Client
- Redirect URI:
http://localhost:8080/callback(or your custom port)
- Save and copy the Client ID (consumer key)
安装
npx @suiteinsider/netsuite-mcp@latest配置
{
"mcpServers": {
"netsuite": {
"command": "npx",
"args": ["@suiteinsider/netsuite-mcp@latest"],
"env": {
"NETSUITE_ACCOUNT_ID": "your-account-id",
"NETSUITE_CLIENT_ID": "your-client-id",
"OAUTH_CALLBACK_PORT": "8080"
}
}
}
}