dorylab/dory

What it solves

AI agents are increasingly capable of writing SQL and exploring database schemas, but their output is often trapped in transient chat logs. Dory solves this by providing a persistent, human-readable workspace where agent-generated SQL, query results, and analytical context can be reviewed, edited, and refined by humans.

How it works

Dory acts as both a traditional SQL client and an MCP (Model Context Protocol) server. When an agent uses Dory, it performs database actions that are rendered as interactive, editable tabs within the Dory interface. This allows users to pick up exactly where an agent left off, visualize data with charts, and maintain a history of database interactions that persists beyond a single chat session.

Who it’s for

This tool is designed for data engineers, analysts, and developers who want to integrate AI agents into their database workflows without losing control or visibility over the executed queries and resulting data.

Highlights

• Editable Agent Workspaces: Turns agent-generated SQL into persistent, modifiable tabs with charts and result sets.
• MCP Integration: Connects seamlessly with agents like Claude Code and Codex CLI to provide a robust database execution layer.
• Schema Explorer: Provides an interactive visual map of database tables and relationships for easier navigation.
• Deep ClickHouse Support: Offers specialized monitoring and management tools specifically for ClickHouse clusters.
• Multi-Database Support: Compatible with a wide range of databases including PostgreSQL, MySQL, SQLite, DuckDB, and Snowflake.