MotherDuck MCP Server

Name: MotherDuck MCP Server
Author: MotherDuck

MotherDuck官方

开始使用源码仓库

Lets agents query and analyze data in local DuckDB and MotherDuck with SQL, including local files and S3 sources.

接入配置

选择客户端，复制配置，再按项目文档补齐环境变量。

一键添加到 Cursor 一键添加到 VS Code

claude mcp add motherduck --env MOTHERDUCK_TOKEN='<MOTHERDUCK_TOKEN>' --env AWS_ACCESS_KEY_ID='<AWS_ACCESS_KEY_ID>' --env AWS_SECRET_ACCESS_KEY='<AWS_SECRET_ACCESS_KEY>' --env AWS_SESSION_TOKEN='<AWS_SESSION_TOKEN>' --env DB_PATH='<DB_PATH>' -- uvx mcp-server-motherduck

~/.codex/config.toml

[mcp_servers.motherduck]
command = "uvx"
args = ["mcp-server-motherduck"]
env = { MOTHERDUCK_TOKEN = "<MOTHERDUCK_TOKEN>", AWS_ACCESS_KEY_ID = "<AWS_ACCESS_KEY_ID>", AWS_SECRET_ACCESS_KEY = "<AWS_SECRET_ACCESS_KEY>", AWS_SESSION_TOKEN = "<AWS_SESSION_TOKEN>", DB_PATH = "<DB_PATH>" }

# codex mcp add motherduck --env MOTHERDUCK_TOKEN='<MOTHERDUCK_TOKEN>' --env AWS_ACCESS_KEY_ID='<AWS_ACCESS_KEY_ID>' --env AWS_SECRET_ACCESS_KEY='<AWS_SECRET_ACCESS_KEY>' --env AWS_SESSION_TOKEN='<AWS_SESSION_TOKEN>' --env DB_PATH='<DB_PATH>' -- uvx mcp-server-motherduck

~/.cursor/mcp.json

{
  "mcpServers": {
    "motherduck": {
      "command": "uvx",
      "args": [
        "mcp-server-motherduck"
      ],
      "env": {
        "MOTHERDUCK_TOKEN": "<MOTHERDUCK_TOKEN>",
        "AWS_ACCESS_KEY_ID": "<AWS_ACCESS_KEY_ID>",
        "AWS_SECRET_ACCESS_KEY": "<AWS_SECRET_ACCESS_KEY>",
        "AWS_SESSION_TOKEN": "<AWS_SESSION_TOKEN>",
        "DB_PATH": "<DB_PATH>"
      }
    }
  }
}

.vscode/mcp.json

{
  "servers": {
    "motherduck": {
      "type": "stdio",
      "command": "uvx",
      "args": [
        "mcp-server-motherduck"
      ],
      "env": {
        "MOTHERDUCK_TOKEN": "<MOTHERDUCK_TOKEN>",
        "AWS_ACCESS_KEY_ID": "<AWS_ACCESS_KEY_ID>",
        "AWS_SECRET_ACCESS_KEY": "<AWS_SECRET_ACCESS_KEY>",
        "AWS_SESSION_TOKEN": "<AWS_SESSION_TOKEN>",
        "DB_PATH": "<DB_PATH>"
      }
    }
  }
}

~/.gemini/settings.json

{
  "mcpServers": {
    "motherduck": {
      "command": "uvx",
      "args": [
        "mcp-server-motherduck"
      ],
      "env": {
        "MOTHERDUCK_TOKEN": "<MOTHERDUCK_TOKEN>",
        "AWS_ACCESS_KEY_ID": "<AWS_ACCESS_KEY_ID>",
        "AWS_SECRET_ACCESS_KEY": "<AWS_SECRET_ACCESS_KEY>",
        "AWS_SESSION_TOKEN": "<AWS_SESSION_TOKEN>",
        "DB_PATH": "<DB_PATH>"
      }
    }
  }
}

claude_desktop_config.json

{
  "mcpServers": {
    "motherduck": {
      "command": "uvx",
      "args": [
        "mcp-server-motherduck"
      ],
      "env": {
        "MOTHERDUCK_TOKEN": "<MOTHERDUCK_TOKEN>",
        "AWS_ACCESS_KEY_ID": "<AWS_ACCESS_KEY_ID>",
        "AWS_SECRET_ACCESS_KEY": "<AWS_SECRET_ACCESS_KEY>",
        "AWS_SESSION_TOKEN": "<AWS_SESSION_TOKEN>",
        "DB_PATH": "<DB_PATH>"
      }
    }
  }
}

.trae/mcp.json

{
  "mcpServers": {
    "motherduck": {
      "command": "uvx",
      "args": [
        "mcp-server-motherduck"
      ],
      "env": {
        "MOTHERDUCK_TOKEN": "<MOTHERDUCK_TOKEN>",
        "AWS_ACCESS_KEY_ID": "<AWS_ACCESS_KEY_ID>",
        "AWS_SECRET_ACCESS_KEY": "<AWS_SECRET_ACCESS_KEY>",
        "AWS_SESSION_TOKEN": "<AWS_SESSION_TOKEN>",
        "DB_PATH": "<DB_PATH>"
      }
    }
  }
}

需要的环境变量

MOTHERDUCK_TOKEN
AWS_ACCESS_KEY_ID
AWS_SECRET_ACCESS_KEY
AWS_SESSION_TOKEN
DB_PATH

把占位值换成你自己的密钥；不要把密钥提交到代码仓库。

接入配置

需要的环境变量

相关推荐