MCP Python Installation Guide

1. Python Environment Installation

Recommended Version: Python 3.9.12 or 3.9.13

Option 1: Install via Anaconda (Recommended)

Use Case: For users requiring a complete scientific computing environment

• Download Links:
  • Anaconda Official Download
  • Domestic Mirror Download

Installation Steps:

1. Double-click the installer to run it.
2. Select the "Just Me" installation mode.
3. It is recommended to keep the default installation path (C:\Users\<Username>\Anaconda3).
4. Check "Add Anaconda3 to my PATH environment variable."
5. After installation, verify in the Start Menu:
   • Open Anaconda Prompt.
   • Execute python --version, which should display 3.9.12/3.9.13.

Option 2: Portable Python (No Installation Required)

Use Case: For quick deployment or restricted environments

• Download Link: Portable Python python.zip
• Usage Steps:
  1. Unzip the file to any directory (avoid paths with Chinese characters or spaces).
  2. Configure environment variables:
     • Add the directory containing python.exe (e.g., D:\python\) to the system PATH.
  3. Verify installation: Run python --version in the command line.

---

2. MCP Environment Configuration

2.1 Obtain the MCP Package

1. Download mcp_excel_xxx.zip and unzip it.
   • Obtain the mcp_excel_xxx.zip file (Download Link) and unzip it to a suitable directory.

Note: The installation package already includes Excel and Python programming support.

2. Recommended path: C:\mcp_excel\ or under the user directory.

2.2 Development Environment Configuration

VSCode Configuration

1. Install VSCode: Official Download Link

2. Open the Project:

   code C:\mcp_excel

3. Essential Plugins:

   • Python extension (ms-python.python)

4. Environment Configuration:

   • Press Ctrl+Shift+P → Select Python: Select Interpreter.
   • Choose Python 3.9.x interpreter.

5. PYTHONPATH Setup:

   • Modify .vscode/settings.json:

     {  
       "python.pythonPath": "python",  
       "python.analysis.extraPaths": [  
         "${workspaceFolder}/lib/X64",  
         "${workspaceFolder}"  
       ]  
     }

   Note: For 32-bit systems, replace with lib/win32.

6. Verify Installation:

   • Run example/calendar/calendar_demo.py.
   • There should be no errors, and calendar data should be output.

PyCharm Configuration

1. Create a New Project:

   • Select the unzipped mcp_excel directory.
   • Choose Python 3.9.x as the interpreter.

2. Configure Paths:

   • File → Settings → Project → Python Interpreter.
   • Click the gear icon → Show All... → Add the following paths:

     C:\mcp_excel\lib\X64  
     C:\mcp_excel

3. Run Tests:

   • Right-click calendar_demo.py → Run.
   • Check if the output is normal.

---

3. Common Issues

1. DLL Load Failure:

   • Confirm system architecture (64-bit requires the X64 directory).
   • Install VC++ Runtime: Download Link.

2. Python Version Mismatch:

   conda create -n mcp python=3.9.12  
   conda activate mcp

3. Path Issues:

   • Avoid paths with Chinese characters or spaces.
   • Run the IDE with administrator privileges.

---

4. Verification

# Test in the Python interactive environment  
from mcp.tools import McpCalendar, McpVersion  
calendar = McpCalendar()  
date = calendar.ValueDate('2025-08-11')  
version = McpVersion()  
print("MCP Python version: ", version)  
print("MCP Python calendar works!")  
print(date)

Technical Support: For issues, contact the Meridian team at support@meridian.com.cn.