ClickHouse MCP

The ClickHouse MCP server connects ClickHouse databases to AI assistants, enabling SQL query execution, schema inspection, and data analysis with read-only safety guarantees.

Installation

pip install mcp-clickhouse
# or
uvx mcp-clickhouse

Configuration

{
  "mcpServers": {
    "mcp-clickhouse": {
      "command": "uvx",
      "args": ["mcp-clickhouse"],
      "env": {
        "CLICKHOUSE_HOST": "your-host",
        "CLICKHOUSE_PORT": "8443",
        "CLICKHOUSE_USER": "your-user",
        "CLICKHOUSE_PASSWORD": "your-password",
        "CLICKHOUSE_SECURE": "true",
        "CLICKHOUSE_VERIFY": "true"
      }
    }
  }
}

ClickHouse Cloud

{
  "env": {
    "CLICKHOUSE_HOST": "your-instance.clickhouse.cloud",
    "CLICKHOUSE_USER": "default",
    "CLICKHOUSE_PASSWORD": "your-password"
  }
}

SQL Playground (Demo)

{
  "env": {
    "CLICKHOUSE_HOST": "sql-clickhouse.clickhouse.com",
    "CLICKHOUSE_USER": "demo",
    "CLICKHOUSE_PASSWORD": ""
  }
}

Available Tools

ClickHouse Tools

• run_select_query - Execute SQL queries (readonly=1 for safety)
• list_databases - List all databases
• list_tables - List tables with pagination and filtering

chDB Tools

• run_chdb_select_query - Query using embedded ClickHouse engine

Features

• Read-Only Safety - All queries run with readonly = 1
• Pagination Support - list_tables supports page_token and page_size
• LIKE Filtering - Filter tables with like or not_like patterns
• Column Metadata - Optional detailed column information
• chDB Support - Embedded ClickHouse for local querying
• Health Check - /health endpoint for HTTP/SSE transport

Environment Variables

Usage Example

Claude, list all tables in the analytics database and then
run a query to show the top 10 users by total purchases
in the last 30 days.

Resources

• ClickHouse Documentation
• chDB Documentation