About aps-aecdm-mcp-dotnet
aps-aecdm-mcp-dotnet is a .NET MCP server sample that connects Claude Desktop with the AEC Data Model API and the Viewer. It grew out of an experiment during an Autodesk Platform Accelerators challenge. The project is presented as a first experiment and is licensed under MIT.
Key features
- 01Built with the ModelContextProtocol .NET SDK
- 02PKCE token retrieval through the GetToken tool
- 03Querying hubs, projects and element groups
- 04Filtering elements by category
- 05Rendering and highlighting models in the Viewer
Use cases
- 01Querying element group data in AEC designs using natural language
- 02Browsing hubs and projects from Claude Desktop
- 03Rendering a design in the Viewer and highlighting selected elements
MCP support
This is an MCP server registered in the Claude Desktop configuration file, exposing seven tools such as GetToken, GetHubs and RenderModel to MCP clients.
Configuration
Local server config is the same for every client. Copy it into your client’s MCP config file.
We haven’t collected a package or remote endpoint for this server. Follow the instructions in the project README.
Tools
Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.
Details
From the project READMEaps-aecdm-mcp-dotnet
.NET MCP Server to connect with Claude Desktop, AEC Data Model API and the Viewer.
Introduction
This sample started as an experiment with the new Model Context Protocol brought as a challenge during one of our Autodesk Platform Accelerators. Special thanks to Mirco Bianchini for bringing this challenge and contributing to get to the solutions presented in this repo.
Prerequisites
To make this work, you'll need to:
- Download and Install Claude Desktop
- Clone or download this repo
- Build this project
- Add a reference to the .csproj in the Claude configuration file (developers resource)
DEMO VIDEO HERE
How it works
This sample creates an MCP server using the ModelContextProtocol .NET SDK.
In this scope we added 4 main tools to our server:
- GetToken to obtain a PKCE token that is used in the APS API requests.
- GetHubs to retrieve the hubs using the AEC Data Model API
- GetProjects to retrieve the projects using the AEC Data Model API
- GetElementGroupsByProject to retrieve the ElementGroups using the AEC Data Model API
- GetElementsByElementGroupWithCategoryFilter to retrieve the elements from one ElementGroup using a category filter.
- RenderModel to render one design with the Viewer
- HighLightElements to highlight elements in the Viewer.
With these tools, you can use natural language to query the data from your elementgroups using the AEC Data Model API.
This is a first experiment with this new protocol. Feel free to submit suggestions and collaborate to this repo so we can improve its functionalities.
Setup
Running locally
Clone this project or download it. It's recommended to install GitHub desktop. To clone it via command line, use the following (Terminal on MacOSX/Linux, Git Shell on Windows):
git clone https://github.com/joaomartins-callmejohn/aps-aecdm-mcp-dotnet
Visual Studio (Windows):
Replace client_id with your own key (Single Page application). You can do it directly in the 'Properties/lauchSettings.json' file or through Visual Studio UI under the debug properties.
You'll need to add a reference to your MCP server in the claude_desktop_congif.json file
{
"mcpServers": {
"aecdm": {
"command": "dotnet",
"args": [
"run",
"--project",
"C:\\Users\\...mcp-server-aecdm.csproj",
"--no-build"
]
}
}
}
Further Reading
Troubleshooting
-
Can't find my hub: Provision your APS app in your hub of Forma for Construction
-
If you made changes to the code and want this to be reflected in Claude, you'll need to end the CLaude task before rebuilding the solution.
License
This sample is licensed under the terms of the MIT License. Please see the LICENSE file for full details.
Written by
João Martins in/jpornelas
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.