mcp-snowflake
A Model Context Protocol server that connects to Snowflake databases and executes SQL queries. It provides tools to list databases, schemas, tables, describe tables, execute read-only queries, sample data, analyze table statistics, profile semi-structured columns, and search columns.
README
MCP Snowflake Server
A Model Context Protocol (MCP) server that connects to Snowflake databases and executes SQL queries.
Features
list_databases- Retrieve a list of accessible databaseslist_schemas- Retrieve a list of schemas from a specified databaselist_tables- Retrieve a list of tables and views from a specified database and schema (supports filtering by name or object type)describe_table- Retrieve detailed structure information for a specified tableexecute_query- Execute read-only SQL queries and return resultssample_table_data- Retrieve sample data from a specified table using Snowflake's SAMPLE ROW clauseanalyze_table_statistics- Generate comprehensive statistical analysis for table columns using approximation functionsprofile_semi_structured_columns- Profile VARIANT/ARRAY/OBJECT columns with sampled flatten-based analysissearch_columns- Search for columns across tables by name pattern and/or data type
Installation
Prerequisites
- Python 3.13 or higher
- uv (Python package manager)
- Access to a Snowflake account
Setup
-
Clone the repository:
git clone <repository-url> cd mcp-snowflake -
Install using uv:
uv tool install -e .
Configuration
Copy the configuration file sample and edit it:
cp .mcp_snowflake.toml.example .mcp_snowflake.toml
[snowflake]
account = "your-account.region"
user = "your-username"
password = "your-password"
warehouse = "your-warehouse"
role = "your-role"
authenticator = "SNOWFLAKE"
Environment variables (SNOWFLAKE__ACCOUNT, SNOWFLAKE__USER, etc.) are also supported.
For full configuration reference including tool toggles, timeout tuning, secondary roles, and browser-based SSO, see docs/settings.md.
Usage
Start the MCP server:
uvx mcp-snowflake --config {your-config-path}
Troubleshooting
Connection Errors
- Verify that configuration file or environment variables are correctly set
- Check that Snowflake account and username are correct
- If
authenticator=SNOWFLAKE, verify password is set correctly - Verify network connectivity
Permission Errors
- Ensure the specified user has permission to access the database
- Set the ROLE if necessary
Development
See docs/development.md.
License
MIT License
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。