ClickHouse

Reviews your ClickHouse setup against 31 production rules, surfacing performance bottlenecks, data-modeling issues, and configuration risks with specific fixes.

Asks about your data volume, query patterns, and latency needs, then recommends a specific table schema, ingestion method, and compression strategy with trade-offs explained.

Scaffolds a working ClickHouse instance on your machine with tables, sample data, and the boilerplate code to start querying. Skips the documentation hunt.

Query a CSV, Parquet, JSON file, S3 bucket, or remote database (Postgres, MySQL, MongoDB, ClickHouse) with SQL — no server setup, no data upload needed.

Swaps your slow pandas queries for a ClickHouse engine underneath — same API, same commands, but runs aggregations and joins on millions of rows in seconds.

Guides you through spinning up a ClickHouse Cloud instance, migrating data from local ClickHouse, configuring access, and validating the production setup end-to-end.

Diagnoses socket hang-ups, Keep-Alive failures, and config mishaps in the @clickhouse/client library. Maps the error to a root cause and a tested fix.

Generates idiomatic queries, inserts, and client setup using the ClickHouse Node.js library. Handles JSON ingestion, result parsing, and connection pooling patterns.

MUST USE when investigating performance issues on a ClickHouse-managed Postgres instance. Provides an evidence-based RCA workflow that scrapes the Prometheus endpoint for system signal, pulls per-digest evidence from the Slow Query Patterns API, and recommends (does not apply) a fix.