Airtable MCP

What is Airtable MCP?

A Model Context Protocol (MCP) server that provides full CRUD operations, schema management, record comments, webhooks, batch operations, governance controls, and AI‑powered analytics for Airtable. It is designed for AI agents and developers who want to interact with Airtable data using natural language through MCP‑compatible clients like Claude Desktop or Claude Code.

How to use Airtable MCP?

Add the server configuration to your Claude Desktop or Claude Code MCP settings using npx, or run the one‑command setup script (for Claude Code). Set the AIRTABLE_TOKEN environment variable with a Personal Access Token (and optionally AIRTABLE_BASE_ID). Once connected, you can issue natural‑language commands such as “List all my bases” or “Create a new task with priority High”.

Key features of Airtable MCP

• Full CRUD on records with filtering and pagination
• Record comments – list, create, update, delete
• Schema management – create and modify tables, fields, and views
• Batch operations – process up to 10 records per operation
• Webhook management – set up real‑time notifications
• Governance & compliance – allow‑list governance, PII masking, exception tracking
• AI analytics – 10 prompt templates for predictive analytics and insights
• Multi‑base support – discover and work with multiple bases dynamically

Use cases of Airtable MCP

• Query, create, update, and delete Airtable records via natural language
• Manage base schemas – add tables, fields, and views programmatically
• Automate record updates with batch create/update/delete operations
• Monitor data changes with webhooks and review notifications
• Run AI‑assisted analytics and generate reports from Airtable data

FAQ from Airtable MCP

What prerequisites are needed to use this server?

Node.js 18 or later, an Airtable account with a Personal Access Token, and optionally a Base ID (which can be discovered via the list_bases tool).

How do I obtain an Airtable Personal Access Token?

Create a token at airtable.com/create/tokens and grant the required scopes: data.records:read, data.records:write, data.recordComments:read, data.recordComments:write, schema.bases:read, schema.bases:write, user.email:read, and optionally webhook:manage.

Can I use the server without specifying a Base ID?

Yes. The Base ID is optional; you can omit it and use the list_bases tool to discover bases dynamically.

Does it support batch operations?

Yes. Three batch tools (batch_create_records, batch_update_records, batch_delete_records) each process up to 10 records per operation for improved performance.

What AI capabilities does it include?

Ten AI prompt templates covering predictive analytics, natural language querying, data quality audits, workflow optimization, and more. These are registered as MCP prompts and can be invoked by AI assistants.