MCP Server Framework

MCP Server Framework

A general-purpose MCP server that provides utility tools for echo, date/time, and file operations within Claude Code and Claude Desktop. It functions as an extensible framework designed to help developers easily build and register custom Python-based tools.

Category
访问服务器

README

MCP Server Framework

A general-purpose Model Context Protocol (MCP) server that provides tools for Claude Code and Claude Desktop.

Quick Start

# Install dependencies
cd mcp-server
pip install -r requirements.txt

# Test the server runs (Ctrl+C to stop)
PYTHONPATH=src python -m mcp_server.server

Project Structure

mcp-server/
├── src/mcp_server/
│   ├── server.py          # Main FastMCP server
│   ├── config.py          # Environment-based configuration
│   └── tools/
│       ├── __init__.py    # Tool registry
│       ├── echo_tool.py   # Example: basic echo tools
│       ├── datetime_tool.py  # Example: date/time utilities
│       └── file_tool.py   # Example: file operations
├── configs/               # Client configuration templates
└── requirements.txt

Configuration

The server is configured via environment variables:

Variable Description Default
MCP_SERVER_NAME Display name for the server mcp-server
MCP_LOG_LEVEL Logging level (DEBUG, INFO, WARNING, ERROR) INFO
MCP_ALLOWED_PATHS Comma-separated paths for file tools (none)
MCP_CUSTOM_VAR_* Custom variables accessible via config (none)

Included Tools

Echo Tools

  • echo - Echo back a message
  • echo_uppercase - Echo in uppercase
  • echo_reverse - Echo reversed

DateTime Tools

  • get_current_time - Get current UTC time
  • get_timestamp - Get current Unix timestamp
  • parse_timestamp - Convert timestamp to readable format
  • time_difference - Calculate time between two timestamps

File Tools (requires MCP_ALLOWED_PATHS)

  • list_directory - List directory contents
  • read_file - Read file contents
  • get_file_info - Get file/directory metadata
  • get_allowed_paths - Show configured allowed paths

Setting Up Clients

Claude Desktop

  1. Open ~/Library/Application Support/Claude/claude_desktop_config.json (macOS)
  2. Add your server configuration:
{
  "mcpServers": {
    "my-mcp-server": {
      "command": "python",
      "args": ["-m", "mcp_server.server"],
      "cwd": "/full/path/to/mcp-server",
      "env": {
        "PYTHONPATH": "/full/path/to/mcp-server/src",
        "MCP_ALLOWED_PATHS": "/Users/you/Documents"
      }
    }
  }
}
  1. Restart Claude Desktop completely (Cmd+Q, then relaunch)

Claude Code

Create .mcp.json in your project root:

{
  "mcpServers": {
    "my-mcp-server": {
      "command": "python",
      "args": ["-m", "mcp_server.server"],
      "cwd": "${workspaceFolder}/mcp-server",
      "env": {
        "PYTHONPATH": "${workspaceFolder}/mcp-server/src",
        "MCP_ALLOWED_PATHS": "${workspaceFolder}"
      }
    }
  }
}

Adding New Tools

  1. Create a new file in src/mcp_server/tools/:
# src/mcp_server/tools/my_tool.py
from typing import TYPE_CHECKING

if TYPE_CHECKING:
    from mcp.server.fastmcp import FastMCP
    from ..config import ServerConfig


def register(mcp: "FastMCP", config: "ServerConfig") -> None:
    """Register my tools with the server."""

    @mcp.tool()
    def my_function(param: str, count: int = 1) -> str:
        """
        Description shown to Claude.

        Args:
            param: What this parameter does
            count: Optional count with default

        Returns:
            What the tool returns
        """
        return f"Result: {param} x {count}"
  1. Register it in src/mcp_server/tools/__init__.py:
def register_all_tools(mcp: "FastMCP", config: "ServerConfig") -> None:
    from . import echo_tool, datetime_tool, file_tool, my_tool  # Add import

    echo_tool.register(mcp, config)
    datetime_tool.register(mcp, config)
    file_tool.register(mcp, config)
    my_tool.register(mcp, config)  # Add registration
  1. Restart Claude Desktop/Code to pick up the new tool.

Tool Design Guidelines

  1. Clear docstrings: The description and parameter docs are sent to Claude
  2. Type hints: All parameters and returns need type hints (defines the JSON schema)
  3. Return strings: Tools should return string results for best compatibility
  4. Error handling: Return user-friendly error messages rather than raising exceptions
  5. Use config: Access config for environment-specific settings (like allowed paths)

Troubleshooting

Server won't start:

  • Ensure PYTHONPATH includes the src directory
  • Check that mcp package is installed: pip install mcp[cli]

Tools not appearing in Claude:

  • Verify the config JSON is valid
  • Check the cwd path is correct
  • Restart Claude Desktop completely (Cmd+Q on Mac)

File tools return "not in allowed directories":

  • Set MCP_ALLOWED_PATHS to comma-separated directory paths
  • Paths must be absolute

Dependencies

  • Python 3.10+
  • mcp[cli]>=1.0.0

推荐服务器

Baidu Map

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。

官方
精选
JavaScript
Playwright MCP Server

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。

官方
精选
TypeScript
Magic Component Platform (MCP)

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。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

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。

官方
精选
本地
TypeScript
VeyraX

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。

官方
精选
本地
graphlit-mcp-server

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。

官方
精选
TypeScript
Kagi MCP Server

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。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

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。

官方
精选