chdb-datastore

Run pandas-compatible tabular workflows on a ClickHouse-backed DataStore.

512|33|Updated Jan 15, 2026
One-click install
npx skills add https://github.com/ClickHouse/agent-skills --skill chdb-datastore
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: chdb-datastore
Source: https://github.com/ClickHouse/agent-skills/tree/main/skills/chdb-datastore
Command: npx skills add https://github.com/ClickHouse/agent-skills --skill chdb-datastore

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

An efficient DataStore for tabular data that accelerates pandas workloads by delegating operations to a ClickHouse-backed engine, reducing runtime and memory pressure.

Core Features & Use Cases

  • Pandas-compatible API: Write familiar pandas-like code that executes as optimized SQL.
  • Cross-source joins: Seamlessly join data from files, SQL databases, and cloud sources.
  • Lazy evaluation: Build up operations and run a single, efficient query when results are needed.
  • Wide data source support: Connect to MySQL, PostgreSQL, S3, MongoDB, Iceberg, Delta Lake, ClickHouse, and more.
  • Use cases: Analyze large datasets, accelerate dashboards, and unify data from multiple sources for analytics.

Quick Start

Compose a DataStore from your tabular data sources and run a simple query to verify results.

Frequently Asked Questions about chdb-datastore

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I speed up pandas data analysis on large datasets?

You can speed up pandas data analysis by delegating tabular workflows to a ClickHouse-backed DataStore, which reduces runtime and memory pressure through lazy evaluation and optimized SQL execution.

Can I join data from different databases and local files in pandas?

Yes, cross-source joins allow you to seamlessly combine tabular data from local files, SQL databases like MySQL and PostgreSQL, and cloud sources like S3 within a single pandas-compatible query.

How does lazy evaluation work for pandas-compatible data queries?

Lazy evaluation builds up operations sequentially and runs a single, efficient query only when results are needed, minimizing memory usage and accelerating complex aggregations and filtering.

Does this pandas acceleration approach support cloud data sources like S3 and MongoDB?

Yes, the DataStore supports 16+ connectors including S3, MongoDB, Iceberg, Delta Lake, ClickHouse, MySQL, and PostgreSQL, enabling efficient inter-source queries across diverse environments.

What is the best way to run cross-source aggregations without running out of memory?

Using a ClickHouse-backed DataStore with a pandas-compatible API and lazy evaluation minimizes memory pressure by pushing filtering, grouping, and aggregation operations to the optimized engine.

Do I need to learn a new API to use ClickHouse for pandas data analysis?

No, you write familiar pandas-like code that the engine automatically translates into optimized SQL, allowing you to accelerate dashboards and unify data without changing your workflow.