Revit MCP Server
xml.Revit.MCP Tools Overview
xml.Revit.MCP provides a comprehensive set of tools for integrating with Autodesk Revit through the Model Context Protocol (MCP). This library serves as a bridge between AI assistants and Revit, enabling powerful automation capabilities and programmatic interaction with building models.

Key Features
The tool library includes numerous functions for Revit automation and interaction:
Basic Operations:
Get available commands from Revit plugin
Execute specified commands in Revit
Call specific Revit functions with parameters
Retrieve view data and selected elements
Element Management:
Find elements by category
Get element parameters and locations
Update element parameters
Delete elements
Show/highlight elements in current view
Move elements to new positions
Creation Tools:
Create levels/floors
Create floor plan views
Create grid lines
Create walls and floors
Create rooms and room tags
Create doors and windows
Create MEP elements (ducts, pipes, cable trays)
Create family instances
Link DWG files
Create sheets
Installation Requirements
xml.Revit: Version 1.3.4.3 or newer
Python: 3.10 or newer
UV Package Manager: Required for installation
Revit: Compatible with versions 2019-2024 (with plugin)
Installation Process
First, install the UV package manager:
pip install uvInstall the revit-mcp package:
pip install revit-mcpTest the installation:
uvx revit-mcpYou should see:
RevitMCPServer - INFO - Successfully connected to Revit on startup
Integration with AI Assistants
Claude for Desktop
Edit claude_desktop_config.json to include:
Cursor
Edit mcp.json to include:
Cline
Edit cline_mcp_setting.json to include:
{
"mcpServers": {
"RevitMCPServer": {
"disabled": false,
"timeout": 30,
"command": "uvx",
"args": ["revit-mcp"],
"transportType": "stdio",
"autoApprove": [
"active_view",
"call_func",
"create_cable_trays",
"create_door_windows",
"create_ducts",
"create_family_instances",
"create_floors",
"create_floor_plan_views",
"create_grids",
"create_levels",
"create_pipes",
"create_room_separation_lines",
"create_room_tags",
"create_rooms",
"create_sheets",
"create_walls",
"delete_elements",
"execute_commands",
"find_elements",
"get_commands",
"get_locations",
"get_selected_elements",
"get_view_data",
"link_dwg_and_activate_view",
"move_elements",
"parameter_elements",
"show_elements",
"update_elements"
]
}
}
}Extending Functionality
You can create custom MCP DLL files to implement additional functionality by:
Implementing the
xml.Revit.MCP.Public.IMCPMethodinterfaceFollowing JSON-RPC 2.0 specification for communication
Compiling to a DLL and placing it in the designated MCP folder
Plugin Configuration
When using the revit-mcp-plugin:
Register the plugin with Revit
Configure commands through: Add-in Modules → Revit MCP Plugin → Settings
Enable the service: Add-in → Revit MCP Plugin → Revit MCP Switch
Once enabled, AI assistants can discover and control your Revit program, executing the various commands provided by the xml.Revit.MCP tools library.
Latest Blog Posts
- Who's Calling? MCP Hosts Are an Identity Blind Spot (And the Spec Knows It)By Om-Shree-0709 on .mcpAgent IdentityOAuth 2.1
- Your AI Chatbot Just Exposed Your CEO's Salary to an InternBy Om-Shree-0709 on .Agent IdentityMCP SecurityOAuth Delegation
- Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)By Om-Shree-0709 on .Agentic AiPrompt InjectionWebAssembly
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/ZedMoster/revit-mcp'
If you have feedback or need assistance with the MCP directory API, please join our Discord server