Smartsheet has launched a generally available Model Context Protocol (MCP) Server that allows AI assistants to read, write, and manage Smartsheet data through a standardized protocol, according to an AWS Machine Learning blog post. The server provides more than 40 tools for exploring and managing sheets, rows, columns, and automations. Supported actions include updating task statuses, creating rows, and modifying project structures through natural-language requests.

The AWS post describes two integration paths. Amazon Quick is positioned for business users, project managers, and operations teams. Through a pre-built Smartsheet connector, users can query project information, update rows, create discussions, and surface risks. Kiro, an agentic integrated development environment, connects to Smartsheet through MCP so developers can reference project specifications, inspect tasks, and track dependencies alongside code. The post also describes a Kiro hook that can update a Smartsheet row when a fix is committed.

The server is available through commercial US, EU, and Australian regional endpoints. It supports Streamable HTTP transport and authentication through OAuth 2.0 or a Smartsheet Personal Access Token. The post says access is governed by existing Smartsheet permissions, with configurable write-tool access and organization-level AI controls. It also states that the MCP server does not store prompts and that Smartsheet receives structured tool calls and their results.

The integrations require appropriate AWS permissions, a qualifying Smartsheet plan, and relevant credentials. Kiro also requires the Kiro IDE and Node.js.