MCP Catalogs
Home

RevitMCPBridge2026

by WeberG619·18·Score 45

RevitMCPBridge2026 provides 705+ MCP endpoints for full AI access to Autodesk Revit via named pipes, enabling direct model manipulation and autonomous goal execution.

developer-toolsai-llmproductivity
7
Forks
0
Open issues
2 mo ago
Last commit
2d ago
Indexed

Overview

RevitMCPBridge2026 is a sophisticated Revit add-in that exposes the entire Revit API through Model Context Protocol using named pipes communication. Unlike HTTP-based alternatives, this implementation uses Revit's own ExternalEvent queue to prevent thread contention and crashes under heavy workloads. The project provides 705+ typed endpoints across 25+ categories, from basic element creation to autonomous goal execution. It includes a 113-file architectural knowledge base covering building codes, standards, and best practices, enabling AI agents to make code-compliant decisions without manual specification.

Try asking AI

After installing, here are 5 things you can ask your AI assistant:

you:Automated generation of construction documents including sheets, views, and annotations from Revit models
you:AI-driven design validation for code compliance and clash detection across multiple disciplines
you:Autonomous creation of building elements based on specifications or requirements documents
you:Why use named pipes instead of HTTP for Revit integration?
you:What levels of AI autonomy are supported?

When to choose this

Choose RevitMCPBridge if you need comprehensive programmatic access to Revit models for AI automation, especially for large-scale modifications, documentation generation, or autonomous goal execution in AEC workflows.

When NOT to choose this

Don't choose this if you're not using Revit 2025/2026, need cross-platform compatibility (Windows-only), or require simpler read-only access to BIM data.

Tools this server exposes

12 tools extracted from the README
  • createWall

    Create a wall with specified parameters

  • createSheet

    Create a new sheet in the Revit model

  • createSheetSet

    Create a complete set of construction document sheets

  • executeGoal

    Execute high-level autonomous goals with parameters

  • getMethods

    List all available methods exposed by the bridge

  • healthCheck

    Check the status of the Revit MCP Bridge connection

  • placeViewOnSheet

    Place a view onto a sheet at specified coordinates

  • configureAutonomy

    Configure autonomy settings, safety guardrails, and allowed methods

  • createRoom

    Create a room with specified parameters

  • getVersion

    Get version information about the Revit MCP Bridge

  • placeDoor

    Place a door in the model at specified location

  • createSchedule

    Create a new schedule with specified fields and filters

Comparable tools

revit-mcpdynamorevit-api-sdkautodesk-bim360

Installation

Installation

Prerequisites

  • Autodesk Revit 2025 or 2026
  • .NET Framework 4.8
  • Visual Studio 2022 (for building from source)

Option 1: Installer Script

git clone https://github.com/WeberG619/RevitMCPBridge2026.git
cd RevitMCPBridge2026
.\scripts\deploy\Install-RevitMCPBridge.ps1

Option 2: Manual Installation

msbuild RevitMCPBridge2026.csproj /p:Configuration=Release
copy bin\Release\RevitMCPBridge2026.dll "%APPDATA%\Autodesk\Revit\Addins\2026%"
copy RevitMCPBridge2026.addin "%APPDATA%\Autodesk\Revit\Addins\2026%"
copy appsettings.json "%APPDATA%\Autodesk\Revit\Addins\2026%"

Claude Desktop Configuration

Add to your Claude Desktop configuration:

{
  "mcpServers": {
    "revit-mcp": {
      "command": "python",
      "args": ["path/to/python-wrapper.py"],
      "env": {
        "REVIT_PIPE_NAME": "\\\\.\\\\pipe\\\\RevitMCPBridge2026"
      }
    }
  }
}

FAQ

Why use named pipes instead of HTTP for Revit integration?
Revit is single-threaded and requires all API calls to execute on the main UI thread. HTTP servers would compete for this thread, causing timeouts, dropped connections, and crashes under load. Named pipes use Revit's own ExternalEvent queue, eliminating thread contention and ensuring stability even with many concurrent requests.
What levels of AI autonomy are supported?
The bridge supports 5 levels of autonomy: 1) Basic Bridge (direct API translation), 2) Context Awareness (tracks element relationships), 3) Learning & Memory (stores corrections), 4) Proactive Intelligence (detects workflow gaps), and 5) Full Autonomy (executes high-level goals with self-healing and guardrails).

Compare RevitMCPBridge2026 with

GitHub →

Last updated · Auto-generated from public README + GitHub signals.