"Just put the MCP server in the repository and share it." That's what the official documentation says. If you place .mcp.json ...
Ensure data pipeline reliability and troubleshoot issues fast by using standard Python logging directly in BigQuery UDFs for real-time observability.