AI agent that generates and executes Python code to interact with Airtable using the MCP code execution pattern from Anthropic's blog.
User Input → Claude Sonnet 4 → Python Code → Subprocess → Airtable MCP → Results Key benefit: Process data in code, not in context.
# Airtable MCP server running at http://localhost:8000/mcp # Anthropic API keypip install langchain langchain-anthropic anthropic aiohttpexport ANTHROPIC_API_KEY="your-anthropic-api-key"python cli.pyYou: List all my Airtable bases 🤖 Agent: Thinking... Generated Code (Attempt 1): ============================================================ from airtable_client import airtable_client async with airtable_client("http://localhost:8000/mcp") as client: from servers import airtable bases = await airtable.list_bases() print(json.dumps({"bases": bases})) ============================================================ ✅ Success! Output: ------------------------------------------------------------ {"bases": [{"id": "appXXX", "name": "My Base"}]} ------------------------------------------------------------ "List all my Airtable bases" "Show me tables in base appXXXXXXXXXXXXXX" "Get all active Orders from my CRM base" "Find records containing 'john@example.com'" "Count records in each table of my base" /help - Show help /clear - Clear conversation history /exit - Exit - User request → "Get all active contacts"
- Claude generates code:
records = await airtable.list_records(...) active = [r for r in records if r['fields']['Status'] == 'Active'] print(len(active)) # Only summary to context!
- Execute in subprocess (sandboxed, validated)
- Return results to user