# Baserow MCP for AI Agents MCP > Baserow MCP connects your AI agent directly to no-code databases, giving you full control over structured data. You can list entire database architectures, explore table schemas down to individual fields, and perform any standard CRUD operation—listing, creating, updating, or deleting records—all via natural conversation. ## Overview - **Category:** loved-by-devs - **Price:** Free - **Tags:** no-code, database-schema, rest-api, data-query, row-management ## Description Baserow lets your AI agent handle complex data tasks without forcing you into a graphical user interface. Instead of clicking through multiple tabs just to find a specific project record or update a field value, your agent acts like a dedicated database administrator. You can start by listing all the databases available, then drill down to view every table and its fields. Need to run an analysis? Ask your agent to query rows using filters and pagination instead of building complex views manually. It handles everything from discovering schemas to running full read/write operations. With Vinkius, you connect this Baserow MCP once, giving your AI client access to thousands of tools, letting it act as a powerful data layer for any project. ## Tools ### create_row Creates a new row in a Baserow table by accepting the table ID and a JSON object containing field names and values. ### delete_row Deletes an entire row from a specified Baserow table, requiring both the table ID and the unique row ID. ### get_row Retrieves all field values for one specific row when you provide the target table ID and row ID. ### get_table Fetches detailed metadata about a specific Baserow table, given its unique identifier. ### list_databases Lists all connected Baserow databases, providing their ID, name, and workspace details so you know what data collections exist. ### list_fields Shows the schema for a given table by listing every field (column), its type, and if it's required. ### list_rows Retrieves multiple rows from a table, allowing you to filter results by specific fields or set page sizes for large datasets. ### list_tables Lists every accessible table across the entire Baserow workspace, showing key details like field count and database location. ### list_views Shows all customized views (like Kanban or Gallery) configured for a specific table, including their filters and sorting rules. ### update_row Modifies one or more fields on an existing row by providing the table ID, row ID, and a JSON object of updates. ## Prompt Examples **Prompt:** ``` List all databases I have access to right now. ``` **Response:** ``` **Databases Found:** * **Projects Tracker**: ID `db-123` (Workspace: Core) * **Team Contacts**: ID `db-456` (Workspace: HR) * **Marketing Leads**: ID `db-789` (Workspace: Sales) ``` **Prompt:** ``` I need to update the status of Project X. The row ID is 500, and I want the new status to be 'Completed' and the completion date set today. ``` **Response:** ``` Success! Row ID `500` in the Projects Tracker table has been updated. **Changes Applied:** * `Status`: Completed * `Completion Date`: 2024-10-27 The record is now fully marked as complete. ``` **Prompt:** ``` What are the column names in my 'Tasks' table? I want to see if there’s a field for priority level. ``` **Response:** ``` **Schema for 'Tasks' Table:** * `Task Name` (Text) * `Assigned To` (Select) * `Due Date` (Date) * `Status` (Single Select) * `Priority Level` (Select) * `Description` (Long Text) ``` ## Capabilities ### Discovering database structures List all available databases and tables in the workspace, providing a complete map of your data architecture. ### Inspecting schemas and views View every field (column) within a table, understanding its type, or listing configured filtered views like Kanban boards and calendars. ### Reading specific records Retrieve the details of any single row by providing its unique ID from a specified table. ### Querying filtered data sets Search and list multiple rows, applying custom filters and setting pagination rules to narrow down large datasets. ### Writing new records Create completely new rows in a specified table by providing field names and corresponding values. ### Modifying existing data Update specific fields on an already existing row, changing only the necessary pieces of information. ### Removing data Permanently delete rows from a table after confirming the action. ## Use Cases ### Generating Quarterly Reports from Scattered Data A Product Manager needs to pull all tasks marked 'High Priority' across three different projects. They ask their agent, which uses list_rows and field filtering, to compile the results into a single summary report. ### Onboarding New Team Members A manager needs to add 20 new staff records. Instead of copying data, they ask the agent to use create_row repeatedly for each person, populating all required fields like start date and department. ### Auditing Data Integrity A developer needs to check if any critical project status records are missing. They instruct the agent to run list_tables followed by get_table to validate the existence and structure of required fields. ### Archiving Old Project Data The team has finished a project and needs to remove all associated task rows from the database. The agent uses delete_row, ensuring only records tied to that specific project are removed. ## Benefits - Control the entire dataset lifecycle: Use list_tables to find your collection, then run list_fields to map out every column before you start. - Eliminate complex UI navigation. Instead of manually updating a record, use update_row to change specific fields on existing data points with a single conversational prompt. - Scale your analysis by running filtered queries: The list_rows tool lets you query thousands of records using field-based filters without building custom views. - Build new content quickly. Use create_row whenever you need the agent to populate an entirely new record into a table for your team. - Gain full visibility into data organization by calling list_views, which maps out how your team has already structured and filtered data in Baserow. ## How It Works The bottom line is you connect your AI client via Vinkius, giving it immediate, conversational access to manage complex database operations inside Baserow. 1. Subscribe to this MCP and enter your personal Baserow Database Token. 2. Connect your preferred AI client (like Cursor or Claude) to Vinkius. This authenticates access to the full catalog of tools, including Baserow. 3. Ask your agent to perform a data task—for example, 'List all databases I have available,' and it executes the query using the provided credentials. ## Frequently Asked Questions **How can Baserow MCP help me find the correct data in my database?** The agent makes it easy by first letting you list all databases and tables. You don't have to guess; you just ask your agent, and it maps out exactly what structures are available for querying. **Does Baserow MCP allow me to change data without using the website?** Yes. You can use the agent to perform full CRUD operations—creating new rows, updating existing records, or deleting old ones—all through simple natural language commands. **What if I need to query thousands of rows from Baserow?** You don't run into limits. The agent handles filtering and pagination automatically when you use list_rows, so you get exactly the subset of data you need without bogging down your client. **Can I see what kind of fields a Baserow table has?** Absolutely. You can ask the agent to run list_fields on any table, and it will return the full schema, showing you if columns are dates, numbers, text, or selections. **Is Baserow MCP only for reading data?** No. It's a write-enabled connection. You can create new records using create_row and modify existing ones with update_row.