Work management context for AI agents with Smartsheet MCP on AWS
Amazon Web Services Galina Jordanowa
Smartsheet now has an MCP server so AI agents can update work data through natural language. It plugs into Amazon Quick and Kiro, so teams can act on live project sheets without switching tools.
Based on reporting by Amazon Web Services, Galina Jordanowa — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
Smartsheet has shipped a generally available Model Context Protocol server, and the pitch is simple: let AI agents work on the same project data humans already use. Instead of asking people to bounce between dashboards, chats, and sheets, the server gives assistants a standardized way to read, write, and manage Smartsheet data.
The new layer exposes more than 40 tools for sheets, rows, columns, and automations. That means an agent can update task status, create rows, change project structure, and pull context from live data through natural language. AWS is framing this around work management itself, where the jobs are messy and cross-functional: timelines shift, tasks get reassigned, and status has to stay current.
The integration paths are split between two different kinds of users. Amazon Quick is the business-user route. A project manager can ask which sprint tasks are blocked this week, and the chat agent can query Smartsheet, pull the relevant rows, and return owner names and due dates without anyone building a report first. Kiro is the developer route. In that setup, a developer can surface open backend tasks inside the coding environment and even have a hook mark a row complete when the fix is committed.
Smartsheet says the server sends structured context, not raw API dumps. That includes row relationships, column types, cell history, and comments, which should make the payloads smaller and cheaper to process. The company says that matters at scale, especially if hundreds of AI-assisted queries are running across a project portfolio.
The server is available through three regional endpoints, uses Streamable HTTP, and supports OAuth 2.0 or a personal access token. Smartsheet also points to enterprise controls: user-level permissions, org-wide AI settings, configurable write access, and no prompt storage on Smartsheet’s side. In other words, the familiar enterprise promise is back: let the agent do more, but only inside the same fences everyone already pretends are comforting.
My take — AI-written commentary, not fact-checked reporting
This is the sort of enterprise AI story that actually makes sense: not a shiny chatbot, just a way to push work around without copy-pasting it into three other systems. The catch is that the real product is still permissions, audit trails, and boring integration glue, which is where most agent hype quietly goes to die. Smartsheet gets points for leaning into that instead of selling magic.
Read more about this at: Amazon Web Services