# Smartsheet MCP server tools The Smartsheet Model Context Protocol (MCP) server provides a way to leverage Smartsheet’s public API via AI clients, allowing you to interact with your data using natural language. The MCP server's underlying **tools** (operations) interact with your Smartsheet data (with your permission) to respond to your prompts. Here are a couple of examples: - **Prompt:** *Summarize what changed in the Marketing Campaign sheet since Monday.* `Tools: get_sheet_summary - Filters on rows where Modified Date > Monday` **Returns:** (An instant change log) - **Prompt:** *What's our biggest bottleneck right now? Look at what's blocking the most items across my sheets.* `Tools: search & get_sheet_summary (multiple times) - AI identifies patterns in blocker columns` **Returns:** (Root cause identification across portfolio) Here is a tool overview and guidance on getting tool details. See also Smartsheet integration with Claude introduces the connector and its capabilities, and provides instructions on getting started using the connector. ## Tool overview The following sections describe some key tools available to your AI client after you've [connected it to the Smartsheet MCP server](/api/smartsheet/ai-integration/install-the-smartsheet-mcp-server). List all the tools To view the most current list of available tools, ask your AI client. **Prompt:** *List the Smartsheet Connector MCP server tools.* Note The tools are subject to Smartsheet API [rate limiting](/api/smartsheet/guides/advanced-topics/scalability-options#rate-limiting) and [standard limitations](/api/smartsheet/guides/basics/limitations). ### Discovery & navigation - **search** - Searches across all accessible Smartsheet assets (sheets, reports, dashboards, folders, workspaces). - **list_workspaces** - Retrieves a list of workspaces with pagination support. - **browse_workspace** - Browses all children (folders, sheets, reports, dashboards) within a workspace. - **browse_folder** - Browses all children (nested folders, sheets, reports, dashboards) within a folder. ### Data retrieval - **get_sheet_summary** - Retrieves sheet data with multi-column filtering, sorting, and column selection. Use this tool to view, analyze, or work with Smartsheet data. * **Best for targeted queries:** Applies filters and column selection to get exactly what you need. * **Filters use AND logic only** (OR isn't supported). * **Large sheets:** Responses are sampled to ~8,000 tokens when sheets exceed this size. *Call `get_columns` FIRST when using filters, sorting, or columns parameters.* Why? Column names are case-sensitive, and `get_columns` gives you the exact names, types (`TEXT_NUMBER`, `DATE`, `PICKLIST`, and more), and available options. *Just exploring?* Call `get_sheet_summary` directly without filters - it returns `available_columns` you can use for follow-up queries. *Tokens* are the basic units that AI language models use to process text. They're chunks of text that can be individual words, parts of words, or even punctuation marks. Both your input and the AI's responses are measured in tokens. This matters because: - Models have maximum token limits (context windows) - Processing speed decreases with increasing token count - **get_sheet_version** - Gets the current version number of a sheet for change detection. - **get_columns** - Retrieves detailed column metadata from a sheet. - **get_cell_history** - Retrieves complete modification history for a specific cell. ### Sheet creation - **create_sheet_in_workspace** - Creates a sheet within a workspace. - **create_sheet_in_folder** - Creates a blank sheet within a folder. - **create_sheet_in_folder_from_template** - Creates a sheet from a template within a folder. ### Row operations - **add_rows** - Adds one or more new rows to a sheet. - **update_rows** - Updates one or more rows in a sheet. (⚠️ destructive) - **delete_rows** - Deletes one or more rows from a sheet. (⚠️ destructive, irreversible) > **Maximum 10 rows** - For safety, you can delete up to 10 rows. ### Column operations - **add_columns** - Adds one or more new columns to a sheet. - **get_columns** - Retrieves detailed column metadata from a sheet. - **update_column** - Updates an existing column's properties. (⚠️ destructive) - **delete_column** - Deletes a column from a sheet. (⚠️ destructive, irreversible) ### Discussion & comment operations - **list_sheet_discussions** - Retrieves all discussions from a sheet. - **list_row_discussions** - Retrieves discussions for a specific row. - **get_discussion** - Retrieves a discussion, including all of its comment threads and attachments. - **create_discussion_on_row** - Creates a discussion, attached to a specific row. - **create_discussion_on_sheet** - Creates a sheet-level discussion. - **add_comment** - Adds a comment to a discussion thread. - **update_comment** - Updates the text of a comment created by the current user. (⚠️ destructive) - **delete_comment** - Deletes a comment. (⚠️ destructive, irreversible) - **delete_discussion** - Deletes a discussion. (⚠️ destructive, irreversible) ### Dashboard operations - **list_dashboards** - Returns a list of dashboards accessible to the current user. - **get_dashboard** - Gets data about a dashboard. ### Report operations - **list_reports** - Returns a list of reports accessible to the current user. - **get_report** - Gets data about a report. ### Workspace & folder operations - **list_workspaces** - Retrieves a list of workspaces. - **browse_workspace** - Browses all children within a workspace. - **create_workspace** - Creates a new workspace. - **create_folder_in_workspace** - Creates a folder in a workspace. - **browse_folder** - Browses all children within a folder. - **create_folder_in_folder** - Creates a nested folder within a folder. ## Get tool details The Smartsheet MCP server is **fully self-describing**, providing built-in tool definitions, detailed parameter schemas, and pre-defined prompt templates. These integrated resources streamline the interaction process, ensuring AI clients can accurately interpret and execute commands with minimal configuration. Chatting with your AI client is the easiest way to learn about the MCP server and get tool details. Here are some example prompts. **Learn the MCP server's capabilities:** *What can I do with the Smartsheet MCP server?* **List underlying tools:** *List all available Smartsheet tools.* **Get a tool description:** *Describe the get_sheet_summary tool.* **Get examples:** *Give me examples of using create_sheet_in_workspace.* **Discover underlying endpoints:** *Which API endpoints does add_rows use?* Tip If you've connected your AI client to multiple MCP servers that share similar tool names, use the `Smartsheet:` prefix to avoid ambiguity. For example, `Describe Smartsheet:list_workspaces` Now that you have the big picture on the tools and know how to get their details, you're ready to operate on project data. ## Related content Smartsheet integration with Claude