airtable-plus-mcp-server
Extends Airtable MCP server with full CRUD, bulk import/upsert, data analysis, schema validation, and data export capabilities.
README
airtable-plus-mcp-server
An extended Airtable MCP server for Claude Code (and other MCP clients). Provides full CRUD plus bulk import/upsert, data analysis, schema validation against known standards, and data export.
Built on domdomegg/airtable-mcp-server (MIT license) — all original tools preserved, with 6 additional power-user tools added.
Quick Start
1. Create an Airtable Personal Access Token
Go to airtable.com/create/tokens and create a token with these scopes:
schema.bases:readschema.bases:writedata.records:readdata.records:writedata.recordComments:readdata.recordComments:write
Set access to the bases you need (or "All current and future bases").
2. Configure Claude Code
Add to ~/.claude.json under mcpServers:
{
"mcpServers": {
"airtable": {
"type": "stdio",
"command": "node",
"args": ["/path/to/airtable-plus-mcp-server/dist/main.js"],
"env": {
"AIRTABLE_API_KEY": "patYOUR_TOKEN_HERE"
}
}
}
}
3. HTTP Transport (optional)
For reusable HTTP mode:
AIRTABLE_API_KEY=patYOUR_TOKEN MCP_TRANSPORT=http PORT=3000 node dist/main.js
Then point any MCP client at http://localhost:3000/mcp.
All Tools (22)
Core Read (6 — from domdomegg)
| Tool | Description |
|---|---|
list_bases |
List all accessible bases |
list_tables |
List tables in a base (with detail level control) |
describe_table |
Get detailed info about a specific table |
list_records |
List/filter/sort records from a table |
search_records |
Full-text search across text fields |
get_record |
Get a single record by ID |
Core Write (3 — from domdomegg)
| Tool | Description |
|---|---|
create_record |
Create a single record |
update_records |
Update one or more records |
delete_records |
Delete one or more records |
Schema Management (5 — 4 from domdomegg + 1 new)
| Tool | Description |
|---|---|
create_table |
Create a new table with field definitions |
update_table |
Update table name/description |
create_field |
Add a field to an existing table |
update_field |
Update field name/description |
delete_field |
NEW — Delete a field from a table |
Comments & Attachments (3 — from domdomegg)
| Tool | Description |
|---|---|
create_comment |
Add a comment to a record |
list_comments |
List comments on a record |
upload_attachment |
Upload an attachment to a record field |
Extended Tools (5 — all new)
| Tool | Description |
|---|---|
bulk_import |
Import up to 1000 records with batching (groups of 10) and rate limiting. Handles errors per-batch with optional stop-on-error. |
bulk_upsert |
Create-or-update records based on merge field(s). Batched with rate limiting. Ideal for syncing data from external sources. |
summarise_table |
Analyse a table: record count, per-field completeness, value distributions, numeric stats (min/max/mean/sum), unique value counts. |
export_table |
Export table data as JSON or CSV. Supports filtering, field selection, sorting. |
validate_schema |
Validate a base against known patterns. Built-in patterns: open_referral_uk (HSDS 3.0), simple_crm, project_tracker. Also accepts custom patterns as JSON. |
Schema Validation Patterns
The validate_schema tool includes three built-in patterns:
open_referral_uk — The Open Referral UK / HSDS 3.0 standard for service directories. Checks for tables like organizations, services, locations, contacts, phones, physical_addresses, regular_schedules, eligibility, taxonomies, and service_taxonomies with their required and optional fields.
simple_crm — A basic CRM pattern with contacts, organisations, interactions, and activities tables.
project_tracker — Project management with projects, tasks, milestones, and team_members tables.
You can also pass a custom pattern as a JSON string:
{
"name": "My Standard",
"description": "Custom schema pattern",
"tables": {
"people": {
"requiredFields": ["name", "email"],
"optionalFields": ["phone", "role"]
}
}
}
Building from Source
git clone <this-repo>
cd airtable-plus-mcp-server
npm install
npm run build
Acknowledgements
This project extends domdomegg/airtable-mcp-server (MIT license). All original tools, types, and architecture are preserved. The extended tools follow the same patterns and conventions.
License
MIT — see LICENSE.
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。