NetSuite AI Connector Service: MCP Server for External AI Clients
NetSuite now exposes an MCP (Model Context Protocol) server that lets external AI clients interact with NetSuite data. A bundled SuiteApp provides standard tools, and developers can build custom MCP tools and MCP Apps with interactive UIs.
What changed
NetSuite introduced the AI Connector Service, an MCP-compliant server that allows external AI clients (assistants, agents, or any MCP-compatible application) to interact with NetSuite data in a standardized way. This is a significant integration surface — MCP is the emerging open protocol for connecting AI models to external data systems, and NetSuite is now a first-party participant.
Key components:
- MCP Standard Tools SuiteApp — A bundled SuiteApp that ships a pre-built set of MCP tools for common NetSuite data operations. This is the fastest path to wiring an AI client into a NetSuite account.
- Custom Tools — Developers can author their own MCP-compliant tools, extending the connector surface beyond the standard set. Custom tools can also support MCP Apps, which expose interactive user interfaces in compatible AI clients (not just text-in/text-out).
- Companion SuiteApp — A separate companion SuiteApp is referenced in the documentation, presumably handling authentication, configuration, or administrative controls for the connector.
What is unclear
The source page is an overview/index and does not specify:
- The exact NetSuite release version this feature ships in or became GA.
- Which standard tools are included in the MCP Standard Tools SuiteApp (record CRUD, saved search execution, SuiteQL, etc.).
- Authentication and authorization model — how tokens/credentials are scoped, whether it leverages OAuth 2.0 / TBA, or introduces a new auth flow.
- Governance or rate-limiting implications for MCP tool invocations.
- Whether custom MCP tools are built with SuiteScript 2.x modules or a separate SDK.
- SuiteApp bundle IDs for the Standard Tools and Companion SuiteApps.
Oracle's sub-pages (linked from this index) likely contain these details. Review them directly:
- Get Started with the NetSuite AI Connector Service
- Creating Custom Tools for the NetSuite AI Connector Service
- Associated Risks, Controls, and Mitigation Strategies
What to do
- Review the sub-topic documentation linked from this page to understand the full setup requirements, including the Companion SuiteApp installation and any role/permission prerequisites.
- Install the MCP Standard Tools SuiteApp from the SuiteApp Marketplace if you want out-of-the-box AI client connectivity. Identify the bundle ID and verify compatibility with your account's release.
- Evaluate the risk and controls page before enabling the service. Exposing NetSuite data to external AI clients introduces data-leakage and prompt-injection risks. Understand what guardrails Oracle provides and what you must configure.
- Plan custom tool development if the standard tools do not cover your workflows. Determine whether custom tools are SuiteScript-based or use a separate tooling framework.
- Terminology note: Oracle uses “AI clients” as a blanket term for assistants, agents, and any MCP-connected application. Other vendors may use different terms — keep this in mind when mapping NetSuite docs to your AI platform’s terminology.
Source: Oracle NetSuite Release Notes