airtable -oauth
Airtable MCP Server With HTTP Streaming & Oauth2.1 Support
claude mcp add --transport stdio onimsha-airtable-mcp-server-oauth uvx airtable-mcp-server-oauth
How to use
This MCP server provides a production-ready interface to interact with Airtable via OAuth 2.0 authentication. Once running, clients can authenticate through the OAuth flow and then perform Airtable operations through the MCP tool set. The server exposes a suite of 10 MCP tools covering core Airtable actions such as listing bases, listing tables, describing table schemas, and performing full CRUD operations on records with support for filtering, sorting, and field selection. Transport is supported via STDIO and HTTP, enabling both local testing and remote integration. After authenticating, you can query the tools to list bases, view and describe table structures, fetch records, create or update records (including batch operations), delete records, and perform advanced searches using Airtable-compatible formulas. The tools are typed, providing clear parameter structures for base_id, table_id, and field data, which improves IDE support and reduces runtime errors.
How to install
Prerequisites:
- Python 3.11+ installed
- uv (the Python package manager) available
- Airtable Developer account to configure OAuth credentials
Installation steps:
- Clone the repository:
git clone https://github.com/onimsha/airtable-mcp-server-oauth.git
cd airtable-mcp-server-oauth
- Install dependencies via uv:
uv sync
- Configure environment variables from the template:
cp .env.example .env
Edit the .env file with your Airtable OAuth credentials and server settings (AIRTABLE_CLIENT_ID, AIRTABLE_CLIENT_SECRET, AIRTABLE_REDIRECT_URI, HOST, PORT, LOG_LEVEL).
- Run the server (HTTP transport example):
uv run python -m airtable_mcp http
You can also run the STDIO transport by omitting the http subcommand:
uv run python -m airtable_mcp
Optional testing:
- Validate with the MCP Inspector tool as described in the Quick Start section of the README.
Note: The project uses uv for dependency management and execution, and the server module is airtable_mcp with an optional http transport. For production deployments, configure environment variables for secure credential handling and adjust HOST/PORT as needed.
Additional notes
Tips and common considerations:
- Ensure AIRTABLE_CLIENT_ID, AIRTABLE_CLIENT_SECRET, and AIRTABLE_REDIRECT_URI are correctly set in .env before starting the server.
- The HTTP transport exposes OAuth endpoints; ensure the redirect URI is registered in your Airtable OAuth application settings.
- The tools are typed; use the exact parameter names (base_id, table_id, etc.) to avoid runtime errors.
- Logs can be tuned with LOG_LEVEL (e.g., INFO, DEBUG).
- If you encounter network or OAuth issues, verify that the redirect URI is reachable from the client performing OAuth.
- Use the MCP Inspector to test endpoints and validate OAuth flow integration before embedding into production clients.
Related MCP Servers
mysql_mcp_server_pro
Model Context Protocol (MCP) server that supports secure interaction with MySQL databases and has anomaly analysis capabilities.更加牛逼!更加好用!不仅止于mysql的增删改查功能; 还包含了数据库异常分析能力;且便于开发者们进行个性化的工具扩展
nautex
MCP server for guiding Coding Agents via end-to-end requirements to implementation plan pipeline
python
🔐 Plug-and-play auth for Python MCP servers.
mcp-yfinance
Real-time stock API with Python, MCP server example, yfinance stock analysis dashboard
yandex-tracker
Yandex Tracker MCP Server with OAuth2 support
cloudwatch-logs
MCP server from serkanh/cloudwatch-logs-mcp