sqlfluff-mcp-server
MCP server that exposes SQLFluff's linting, fixing, and parsing capabilities, enabling MCP-aware clients to lint and fix SQL directly.
README
sqlfluff-mcp-server
An MCP server that exposes SQLFluff's linting, fixing, and parsing over the Model Context Protocol, so any MCP-aware client (Claude, other agents, IDE integrations) can lint and fix SQL directly.
This project is independent of the SQLFluff maintainers — it's a thin wrapper
around the public sqlfluff Python package.
Tools
| Tool | Input | Config resolution |
|---|---|---|
lint_file |
path on disk | walks up from the file for .sqlfluff / pyproject.toml / etc. |
fix_file |
path on disk (+ write flag) |
same directory walk |
parse_file |
path on disk | same directory walk |
lint_sql |
raw SQL text + dialect |
none — dialect supplied explicitly |
fix_sql |
raw SQL text + dialect |
none — dialect supplied explicitly |
parse_sql |
raw SQL text + dialect |
none — dialect supplied explicitly |
list_dialects |
— | lists supported dialect names |
The *_file tools honor project config the same way the sqlfluff CLI does
(via FluffConfig.from_path, which walks up the directory tree looking for
.sqlfluff, pyproject.toml, setup.cfg, or tox.ini). The *_sql tools
are for content that hasn't been written to disk (e.g. streamed from an
editor buffer) and require you to pass dialect explicitly since there's no
file location to resolve config from.
Requirements
- Python 3.10+
- uv (recommended) or
pip
Install
git clone <this-repo-url>
cd sqlfluff-mcp-server
uv sync # or: pip install -e ".[dev]"
Run
uv run sqlfluff-mcp-server # or: sqlfluff-mcp-server, if installed on PATH
This starts the server over stdio, the standard transport for MCP clients that launch servers as a subprocess (Claude Desktop, Claude Code, etc.).
Registering with an MCP client
Example Claude Desktop / Claude Code config entry:
{
"mcpServers": {
"sqlfluff": {
"command": "uv",
"args": ["--directory", "/path/to/sqlfluff-mcp-server", "run", "sqlfluff-mcp-server"]
}
}
}
Development
uv sync --extra dev
uv run pytest
uv run ruff check .
Notes on dependencies
mcpis pinned to<2.0.0. The SDK's 2.0 line restructured the high-level server API (no moremcp.server.fastmcp.FastMCP); this project hasn't been migrated yet.sqlfluffis left unpinned above3.0.0— SQLFluff releases fairly often and this server only depends on its stableLinter/FluffConfigAPI.
License
MIT — see 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。