# SeaTable MCP for AI Agents AI Agent Connect > SeaTable MCP for AI agents. Connect your database to your agent to manage records, run SQL queries, and handle complex data tasks with natural language. It's built for project management, CRM tracking, and operational workflows that need more power than a basic spreadsheet but less friction than a traditional database. You can update statuses, pull reports, and manage your entire workspace without leaving your chat window. ## Overview - **Category:** industry-titans - **Price:** Free - **Endpoint:** https://edge.vinkius.com/vk_preview_EcrX5RuTBwMiiV5mY6fQqN7e2NIAfxegGUxfqa1y/ai-agent-connect - **Tags:** seatable, database-api, collaborative-data, sql-query, spreadsheet-database, no-code-automation, data-orchestration, mcp ## Description SeaTable combines the power of a professional database with the ease of use of a spreadsheet. This integration lets you take full control of your collaborative workflows through natural conversation. You can pull row metadata, add new records, and run complex SQL queries right from your chat interface. Instead of jumping between tabs to find a specific status or manually typing in new entries, you just tell your agent what to do. Vinkius makes it easy to get this connected so your AI acts as a dedicated database administrator. It handles the messy stuff like swapping permanent tokens for short-lived access keys to keep your data safe. You get a workspace that moves as fast as you can type. You can manage your CRM, track project milestones, or monitor asset logs without ever opening a new window. It bridges the gap between your high-level plans and the actual data that supports them, allowing you to stay in your flow state while the agent handles the heavy lifting of data entry and retrieval. Whether you're cleaning up a messy inventory list or pulling a custom report for a stakeholder, this connection ensures your data remains the single source of truth. ## Tools ### get_row Fetches a single specific record from a table when you need to see one item. It's the fastest way to grab a specific piece of information. ### list_columns Shows all the headers and data types in a table to understand the schema. This helps you know exactly what data points are available. ### list_rows Pulls every record currently in a table for a full overview of your data. You can use this to see everything in one go. ### list_tables Provides a full list of tables and their columns for a broad workspace view. It helps you navigate your entire database structure. ### list_views Shows all the different views you've set up for a table to see different perspectives. This is useful for seeing how others are looking at the data. ### query_sql Runs standard SQL queries against your database for advanced filtering and aggregation. You can perform complex data analysis using familiar commands. ### delete_row Removes a specific record from a table. This helps keep your workspace clean by getting rid of outdated information. ### get_base_metadata Pulls the underlying structure and information for your current base. It gives you a clear look at how your data is organized. ### create_row Adds a new entry to a table using a JSON string for easy data input. You can quickly populate your database without manual typing. ### create_table Builds a new table within your workspace to organize new categories of data. This allows you to expand your database structure on the fly. ### update_row Modifies the data in an existing record to keep your project information current. Use this to change status updates or new details in a row. ## Prompt Examples **Prompt:** ``` List all rows from the 'Inventory' table in SeaTable. ``` **Response:** ``` I've retrieved the records. You have 15 items in the 'Inventory' table, including **Laptop Pro** (Stock: 5) and **Office Chair** (Stock: 12). Would you like me to filter for low stock? ``` **Prompt:** ``` Show me all tables in the project database and pull the data from the Tasks table with filters. ``` **Response:** ``` Project database has 6 tables. * **Tasks** (234 rows): 45 open, 123 in progress, 66 completed. * **Team Members** (18 rows) * **Milestones** (12 rows, 3 overdue) * **Budget** (45 rows, total $234K) * **Documents** (89 attachments) * **Time Logs** (456 entries, 1,234 hours) Tasks filtered by status=open: 45 tasks. Top priority: **API redesign** (due tomorrow, assigned to Sarah). 8 tasks are blocked. Average completion time: 4.2 days. Overdue: 12 tasks. ``` **Prompt:** ``` Create a new table called Sprint Backlog with columns for story points, assignee, and sprint number. ``` **Response:** ``` Table **Sprint Backlog** created successfully. **Columns configured:** * **Task Name** (text, primary) * **Description** (long text) * **Story Points** (number, range 1-13) * **Assignee** (collaborator) * **Sprint Number** (single select: Sprint 1-10) * **Status** (single select: To Do, In Progress, Review, Done) * **Priority** (single select: Critical, High, Medium, Low) * **Created Date** (date, auto-fill) Default view: grouped by Sprint Number. 3 additional views created: Kanban by Status, Calendar view, and My Tasks filter. ``` ## Capabilities ### Update records automatically Modify existing data points in your database using simple natural language commands. ### Run SQL queries in chat Execute complex filters and data aggregations using standard SQL syntax. ### Sync data across tables Keep your collaborative data consistent by programmatically updating rows. ### Manage database metadata Retrieve the underlying structure and information for your entire workspace. ### Monitor system activity Track database records and system status through direct AI interaction. ### Build new tables via chat Create new data structures and columns on the fly to organize new categories. ## Use Cases ### Inventory Management An inventory manager asks their agent to list all items with stock below 5. The agent uses list_rows and filters the data to highlight urgent needs. ### CRM Data Cleanup A sales lead wants to update the status of 50 leads. They ask the agent to use update_row to change 'Pending' to 'Contacted' for a specific list. ### Project Milestone Tracking A project manager asks for a summary of all overdue tasks. The agent runs a query_sql command to aggregate the data and presents a clean summary. ### Database Auditing A data engineer needs to see the current schema of the production base. The agent uses get_base_metadata to pull the structure in seconds. ## Benefits - Run complex SQL queries and data aggregations without ever writing a single line of code in your chat client. - Keep your project timelines and task statuses current by using update_row to modify records instantly. - Manage your entire database structure and metadata without having to navigate through nested browser menus. - Create new tables and data categories on the fly using create_table to expand your workspace as your needs change. - Synchronize your collaborative data across various tables to ensure everyone is looking at the same source of truth. - Securely handle API tokens and data access, allowing your agent to perform actions safely on your behalf. ## How It Works The bottom line is you get a database administrator that lives inside your chat interface. 1. Subscribe to the SeaTable MCP on Vinkius. 2. Input your SeaTable API Token and Server URL in your settings. 3. Ask your agent to perform tasks like 'Show me all overdue tasks' or 'Update the status of the API redesign'. ## Frequently Asked Questions **Can I use SeaTable MCP to update my project tasks?** Yes. You can ask your agent to update specific rows, such as changing a task status from 'In Progress' to 'Done' or updating a due date without opening the browser. **How do I run SQL queries on my SeaTable data?** You can simply describe the data you want to see in plain English. Your agent will translate that into a SQL query to pull the specific filters or aggregations you need. **Is SeaTable MCP safe for my company's database?** Yes. The Connector handles the complex security of exchanging permanent tokens for short-lived access tokens, ensuring your data remains secure while your agent performs actions. **Can I create new tables in SeaTable using an AI agent?** Absolutely. You can describe the columns and purpose of a new table, and the agent will build it for you instantly to organize your new data categories. **How does SeaTable MCP help with CRM management?** It allows you to manage your CRM records through conversation. You can pull lead lists, update contact statuses, and query aggregate sales data without manual data entry. **Can SeaTable MCP handle large amounts of data?** Yes. By using SQL queries and row listing, the Connector can efficiently navigate large datasets to pull only the specific information you need at any given time. **Can my AI automatically run SQL queries on my SeaTable data?** Yes! Use the `query_sql` tool. Provide a standard SQL string (e.g., 'SELECT * FROM Tasks WHERE Status = "Done"'), and your agent will return the aggregate or filtered results instantly. **How do I find my SeaTable API Token?** Open your SeaTable base, click the three dots (Advanced) next to the base name, select **API Token**, and create a permanent token for the base. **Does this integration work with self-hosted SeaTable instances?** Yes! You can provide your custom `serverUrl` (e.g., https://seatable.mycompany.com) during setup to connect the Connector to your private instance.