Update 25.10.30 Test In Revit 2021 , 2023 , 2024 , 2025 , 2026

1. Fix The Muti-Version Build Failed : Thats The the net8.0 | net11.0-windows contains the require keyword , but cant use this keyword in <= 10.0 c# version
2. Add Setup Video
   • Change The Version Change The Version
   • Move The Addin File Or Change Target Output Path in .csproj Move Add-In File
   • Run Command (Better Eng) Run Command

Update 25.10.28

1. Fix The Project Than Cant Run Success , Reason : the server path was not auto adaptive to other user

2. ReWrite The RevitTest Project , Also Add The .addin File To User

   • Addin File

3. Add The Function That can read api_key file

4. Other detail Can Read :

   • NET.MCP.CLinet Update
   • Revit Process Update

5. Change Solution To Muti-Version

   • NET.Mcp.Client
   • NET.Mcp.Server
   • RevitTest.csproj

6. Add New Utility File

   • ConvertRevitCommand.cs
   • FileUtility.cs
   • IRevitCommand.cs
   • RequestHandler.cs

7. In Revit 2026 , The Sturucture Was Refernce NewtonSoft 13.0.1 . So If Project Ref. Others Version , Will Throw Error : File Was Loaded .....

8. Also , if use the Revit Add-iN Manage , use this code will cant find location()

   • if use this tool to debug , you need to make FunctionUserCallWindow.xaml.cs Line 131 ~147 the client_path to local machine path .
   • if use this with addin , you can use the fileutility to get lcoation.

9. Add Transaction Method To Functions

Revit MCP Blog

BLOG_ZH | BLOG_EN

Currently, the internet is flooded with various AI-related information. As recent development work involves AI technology, this article introduces the integration method between MCP and Revit to achieve the "one-click modeling" feature in public accounts.

This is just one of the possible implementation approaches, and other technical paths exist. However, this method allows for rapid deployment, helping teams quickly explore relevant application directions.

Function Call

Before discussing MCP, it's essential to introduce Function Calling—a capability provided by LLM APIs. By defining Function Call specifications during interactions with the LLM, the model can autonomously select and execute the appropriate functions when needed. **Picture From：ailydoseofds ** Taking DeepSeek as an example, its API documentation provides a clear guide on how to use Function Calling.

tools = [  
	{  
		"type": "function",  
		"function": {  
			"name": "get_weather",  
			"description": "Get weather of an location, the user shoud supply a location first",  
			"parameters": {  
				"type": "object",  
				"properties": {  
					"location": {  
						"type": "string",  
						"description": "The city and state, e.g. San Francisco, CA",  
					}  
			},  
			"required": ["location"]  
			},  
		}  
	},  
]

The underlying logic is that the LLM, when generating a response, will:

1. Select the appropriate function based on the context and generate compliant parameters.
2. Execute the function.
3. Retrieve the result and return it.
4. Summarize the output and deliver it to the client interface.

MCP

Introduce MCP

MCP (Model Context Protocol) is an extension protocol

... View full README on GitHub