Data Warehouse & Lakehouse MCP Servers — Snowflake, BigQuery, Databricks, ClickHouse, DuckDB, Redshift
At a glance: Data warehousing has exceptional MCP coverage — every major platform has official support. ClickHouse leads the open-source community with 720 stars. BigQuery offers a fully managed re...

Source: DEV Community
At a glance: Data warehousing has exceptional MCP coverage — every major platform has official support. ClickHouse leads the open-source community with 720 stars. BigQuery offers a fully managed remote server. Snowflake brings Cortex AI integration. This is one of the strongest enterprise MCP categories. Rating: 4.5/5. ClickHouse — Official Server (720 stars) Detail Info ClickHouse/mcp-clickhouse 720 stars, Python, 4 tools Most starred data warehouse MCP server. Dual engine: remote ClickHouse clusters + embedded chDB engine. Safety defaults: read-only by default, write requires --allow-write, destructive ops need --allow-destructive. Three transports (stdio, SSE, Streamable HTTP). Limitation: Only 4 tools — no describe_table without writing SQL. DuckDB / MotherDuck — Official Server (441 stars) Detail Info motherduckdb/mcp-server-motherduck 441 stars, Python, MIT, 5 tools Universal data access: local DuckDB files, in-memory, S3, and MotherDuck cloud — all from the same server. switch_d