Airtable MCP. Talk to your spreadsheets and manage structured data via chat.

Your AI agent treats your Airtable data like a natural conversation. You stop writing clunky API calls and just talk to your bases. The agent handles all the querying, filtering, and updating for you. You'll use this to read schemas, find records, update sheets, and build automated workflows right from your chat window.

List available bases and tables: You can list every primary base and every individual table across your entire workspace. The agent reads your full data structure, showing you exactly what you're working with.

Retrieve base schema details: You get a detailed rundown of every column and its data type for any specific table. You can check the field names and types before you write a single query.

Search and filter records: You tell the agent to find specific rows using complex filters and formulas—like finding every record where the 'Status' is 'Done.' The agent runs these formula-driven queries to pull out exactly what you need.

Create and modify records: You can ask the agent to add new entries or change existing data. You'll use create_records to bulk-add data to a table, and update_records to change status or ownership across multiple records at once. You'll also use delete_records to safely remove outdated or duplicate data in bulk.

Read specific records and lists: You can fetch the details of a single record by its ID using get_record, or you can list all records from a table using list_records. For targeted searches, search_records finds records using complex filters, while list_tables shows all the tables inside a single base.

If you need to see everything, you can run list_bases to get all the primary bases in your account. If you're working inside a base, you can use list_fields to see all the column names and types for a specific table.

Use Case Example: You want to know every lead in your CRM that hasn't been touched since last month. You ask your agent to search_records using date criteria. The agent finds the matching rows. Next, you tell it to update the 'Last Contacted' field for all those records to today's date using update_records.

You've just run a complex workflow—query, then modify—without leaving your chat client.