MCP.so
登录

MCP Snowflake Reader

@fnf-deepHeading

基本信息

分类

数据库

许可证

MIT

传输方式

stdio

作者

fnf-deepHeading

关于 MCP Snowflake Reader

A Model Context Protocol (MCP) server that provides secure, read-only access to Snowflake databases. Allows LLMs to safely query tables and describe schemas without write permissions. Built with Python and the official Snowflake SDK.

MCP Snowflake Reader 配置

使用下面的配置,将此服务器添加到你的 MCP 客户端。

{
  "mcpServers": {
    "mcp-snowflake-reader": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "mcp-snowflake-reader",
        "--connection",
        "{\"account\":\"your-account\",\"user\":\"your-user\",\"password\":\"your-password\",\"warehouse\":\"your-warehouse\",\"database\":\"your-database\",\"schema\":\"your-schema\",\"role\":\"your-role\"}"
      ]
    }
  }
}

MCP Snowflake Reader 工具

工具列表暂未提供

当前收录暂时没有可展示的工具列表,不代表服务没有工具。请查阅项目文档了解支持的能力。

工具是从 README 中自动提取的。维护者可以在 ## Tools 标题下列出工具,即可填充这部分内容。

概览

What is MCP Snowflake Reader?

MCP Snowflake Reader is a read-only MCP server that provides secure, read-only access to Snowflake databases through the MCP protocol.

How to use MCP Snowflake Reader?

Configure your MCP client (Cursor AI or Claude) with a Docker or uvx command, passing a JSON connection string with your Snowflake account, user, password, warehouse, database, schema, and role. Alternatively, install automatically via Smithery.

Key features of MCP Snowflake Reader

  • Read-only access to Snowflake databases
  • Connection via JSON configuration string
  • Supports Docker and uvx runtimes
  • Installable via Smithery for Claude Desktop
  • MIT licensed

Use cases of MCP Snowflake Reader

  • Query Snowflake data without write risk from MCP clients
  • Provide safe, read-only database access to AI assistants
  • Integrate Snowflake analytics into Cursor AI or Claude workflows
  • Automate table inspection and reporting in a controlled environment

常见问题

1

Can MCP Snowflake Reader write or modify data?

No, only read-only operations are allowed. SQL keywords like INSERT, UPDATE, DELETE, DROP, ALTER, CREATE, GRANT, REVOKE, COMMIT, and ROLLBACK are explicitly prohibited.

2

What connection credentials are required?

You must provide a JSON object containing account, user, password, warehouse, database, schema, and role.

3

Are there any restrictions on table names?

Table names can only contain alphanumeric characters, underscores, and dots.

4

What runtimes are supported?

The server can be run via Docker or uvx. Configuration is provided in the MCP client settings file.

5

How do I install MCP Snowflake Reader on Claude Desktop automatically?

Run `npx -y @smithery/cli install @fnf-deepHeading/mcp-snowflake-reader --client claude` via Smithery.

评论

数据库 分类下的更多 MCP 服务器