DatusAI

AI-native data agent platform that automates data engineering tasks with intelligent planning and execution.

MCPCLI

About DatusAI

AI-native data agent platform that automates data engineering tasks with intelligent planning and execution. Explore how DatusAI integrates with the agentic data stack ecosystem and supports autonomous data operations.

Key Features

  • AI-native SQL client ('Claude Code for data engineers') with natural language to SQL generation
  • Evolvable context engine capturing metadata, semantic models, metrics, and reference SQL
  • Two execution modes: Agentic Mode for exploratory workflows, Workflow Mode for production stability
  • Subagent system to convert data domains into domain-aware chatbots with scoped context
  • Continuous learning from user feedback and success stories to improve accuracy
  • MCP extensions for integrating external tools and servers
  • Database adapter support for PostgreSQL, MySQL, Snowflake, DuckDB, StarRocks, ClickZetta
  • Three entry points: Datus-CLI (engineers), Datus-Chat (web UI), Datus-API (programmatic access)

Agent Integration

MCP Server

MCP integration available

CLI — datus-agent

$ pip install datus-agent && datus-agent init
CLI Documentation

External Links

Datus-agent GitHub Repository

Main open-source repo: CLI SQL client with AI-native context engineering for data

Datus Official Documentation

Full docs covering CLI, Chat, API components and agent workflow configuration

Datus-ai GitHub Organization

Organization page listing all DatusAI repositories and releases

Datus Blog

Technical blog with posts on AI-native data engineering patterns and case studies

Datus-agent Releases

Release notes showing MCP integration and feature history

← Back to Data Agents