Reg-Kris/mcp-server-py

mcp-server-py

Python MCP (Model Context Protocol) server exposing Airtable tools for LLM integration

Overview

This microservice implements the Model Context Protocol (MCP) to expose Airtable operations as tools that can be used by LLMs like Claude or Gemini. It communicates with the airtable-gateway-py service to perform actual Airtable operations.

Available Tools

1. list_tables - List all tables in a base
2. get_records - Retrieve records from a table
3. create_record - Create a new record
4. update_record - Update an existing record
5. delete_record - Delete a record
6. search_records - Search records with filters
7. create_metadata_table - Create a table with metadata about all tables in a base

Quick Start

# Clone the repository
git clone https://github.com/Reg-Kris/mcp-server-py.git
cd mcp-server-py

# Create virtual environment
python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

# Install dependencies
pip install -r requirements.txt

# Set environment variables
cp .env.example .env
# Edit .env with your configuration

# Run the MCP server
python -m src.server

# Or run with stdio transport
python -m src.server --transport stdio

Testing with MCP Inspector

# Install MCP inspector
npm install -g @modelcontextprotocol/inspector

# Run inspector
mcp-inspector python -m src.server

Environment Variables

AIRTABLE_GATEWAY_URL=http://localhost:8002
AIRTABLE_GATEWAY_API_KEY=simple-api-key
LOG_LEVEL=INFO

Integration

This server can be integrated with:

• Claude Desktop (via claude_desktop_config.json)
• Custom LLM applications
• The llm-orchestrator-py service

Docker

# Build image
docker build -t mcp-server-py .

# Run container
docker run -p 8001:8001 --env-file .env mcp-server-py