DuckDB MCP Server

Stabledatabase
230 starsPythoncommunity
GitHub
About

Overview

The DuckDB MCP Server enables AI agents to use DuckDB, the fast in-process analytical database. It can query CSV, Parquet, JSON, and other file formats directly without loading them into a database first. Ideal for ad-hoc data analysis, ETL pipelines, and working with large datasets locally.
Capabilities

Tools & Capabilities

query

Execute an analytical SQL query using DuckDB

list_tables

List all tables and views

describe_table

Get the schema of a table

import_file

Import a CSV, Parquet, or JSON file as a table

Setup

Installation

bash
Install
pip install mcp-server-duckdb
Examples

Example Usage

javascript
Usage
{
  "mcpServers": {
    "duckdb": {
      "command": "python",
      "args": ["-m", "mcp_server_duckdb"],
      "env": {
        "DUCKDB_PATH": "./analytics.duckdb"
      }
    }
  }
}

Quick Info

Authorcommunity
LanguagePython
StatusStable
Stars 230
Last UpdatedFeb 12, 2026

Need a Custom MCP Server?

Our team builds custom MCP servers tailored to your workflow.

Get in Touch

Need a Custom MCP Server?

Our team builds custom MCP servers tailored to your workflow. From proprietary data sources to internal tools, we have you covered.

Contact Us