What are the runtime dependencies?
Python 3.10+ and Pydantic 2 are required. Flask, Gunicorn, and Redis are optional for web server, production deployment, and response queuing respectively.
@fulfilio
基本信息
Python utilities to work with MCP servers
使用下面的配置,将此服务器添加到你的 MCP 客户端。
{
"mcpServers": {
"mcp-utils": {
"command": "npx",
"args": [
"-y",
"@smithery/cli",
"install",
"mcp-proxy",
"--client",
"claude"
]
}
}
}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。
点击上方按钮,在临时沙箱中运行此服务器并实时获取工具列表。
mcp-utils is a Python utility package for building Model Context Protocol (MCP) servers, focused on synchronous implementations using Flask. It provides helpers, decorators, and Pydantic models for developers integrating MCP into existing Python applications without asynchronous code.
Install with pip install mcp-utils. Create an MCPServer instance, decorate handlers with @mcp.tool() or @mcp.prompt(), and handle incoming requests via mcp.handle_message(). For production, deploy with Flask and Gunicorn.
Python 3.10+ and Pydantic 2 are required. Flask, Gunicorn, and Redis are optional for web server, production deployment, and response queuing respectively.
No, mcp-utils is explicitly designed for synchronous implementations. For async MCP servers, consider the MCP Python SDK.
Claude Desktop does not natively support SSE/HTTP; use `mcp-proxy` (installable via PyPI or Smithery) to bridge stdio with the server's SSE endpoint.
Yes, the README provides Flask examples with Streamable HTTP and optional SQLAlchemy transaction handling, plus a Gunicorn deployment example.
Use the MCP Inspector (`npm install -g @modelcontextprotocol/inspector`) to open a web interface, point it to your server’s SSE URL, and send test messages.
AI Agents & MCPs & AI Workflow Automation • (~400 MCP servers for AI agents) • AI Automation / AI Agent with MCPs • AI Workflows & AI Agents • MCPs for AI Agents
MCP Toolbox for Databases is an open source MCP server for databases.

B2B competitive intelligence, live sales battlecards, 5-axis ICP prospect fit scoring, and tech stack teardowns over MCP and REST with cited public-web evidence.

See and control a real Windows PC you own — from any MCP client, locally or remotely. Observe (UIA + screenshots), click/type/drag/scroll, launch apps, owner Live View. BYOH: your machine, your key.
Hindsight: Agent Memory That Learns
评论