Revit MCP: Read data from Active document

Overview

Revit MCP is a desktop assistant that lets you talk to an Revit model in plain language. Instead of clicking through menus or writing one: off scripts, you describe what you want, and the app decides which tools to call against the active document.

The stack is split across two runtimes on purpose:

• WPF (.NET 8) hosts the UI and Gemini integration
• Revit add-in (.NET Framework 4.8) talks to the Revit DB API inside the Revit main thread.

The bridge between those two is Named Pipe

I have exposed some ( three ) APIs through MCP Server. Those are

• GetWallLength

public object GetWallLength(string wallId)
        {
            var id = new ElementId(long.Parse(wallId));
            var wall = _doc.GetElement(id) as Wall;
            if (wall == null) return new { error = "Wall not found" };

            var lengthInternal = wall.get_Parameter(BuiltInParameter.CURVE_ELEM_LENGTH).AsDouble();
            var lengthMm = UnitUtils.ConvertFromInternalUnits(lengthInternal, UnitTypeId.Millimeters);
            var lengthM = UnitUtils.ConvertFromInternalUnits(lengthInternal, UnitTypeId.Meters);

            return new
            {
                wall_id = wallId,
                length_mm = System.Math.Round(lengthMm, 2),
                length_m = System.Math.Round(lengthM, 3)
            };
        }

• GetFloorArea

public object GetFloorArea(string levelName)
        {
            var level = new FilteredElementCollector(_doc)
                .OfClass(typeof(Level))
                .Cast<Level>()
                .FirstOrDefault(l => l.Name == levelName);

            if (level == null) return new { error = $"Level '{levelName}' not found" };

            var floors = new FilteredElementCollector(_doc)
                .OfClass(typeof(Floor))
                .Cast<Floor>()
                .Where(f => f.LevelId == level.Id)
                .ToList();

            var totalArea = floors
                .Sum(f => f.get_Parameter(BuiltInParameter.HOST_AREA_COMPUTED).AsDouble() * 0.0929);

            return new { level = levelName, area_sqm = System.Math.Round(totalArea, 2) };
        }

• ListRoomsByLevel - which is way easier as you only have to write one LINQ...

var rooms = new FilteredElementCollector(_doc)
                .OfClass(typeof(SpatialElement))
                .OfType<Room>()
                .Where(r => r.Level?.Name == levelName)
                .Select(r => new
                {
                    name = r.Name,
                    number = r.Number,
                    area_sqm = System.Math.Round(r.Area * 0.0929, 2)
                })
                .ToList();

Full Request Flow

User types prompt in WPF UI
        ↓
GeminiService.SendPromptAsync()
        ↓ HTTPS
Google Gemini API (gemini-2.5-flash)
        ↓ tool_use: open_revit_document
McpClientService.CallToolAsync()
        ↓ stdio (MCP protocol)
RevitMCP.Server.exe
        ↓ named pipe (RevitMCPPipe)
RevitMCP.Addin (inside Revit.exe)
        ↓ Revit DB API
Real Revit data (.rvt file)
        ↑ JSON result
Back through pipe → MCP server → Gemini
        ↑ tool_result
Gemini reasons, calls more tools if needed
        ↑ final text answer
ResponseText shown in WPF UI

2 cents and conclusion

It took me 12 days to build this while I was automating Revit workflows. As a new-generation engineer, I thought, "Why not build AI projects?"

They are very simple and easy to do in my openion. I think this is way simpler than automating Revit workflows and writing those LINQs, yet it will still be considered a very good project for a resume. Haha.

This is my first blog post and I hope to write more in the future. let's see if I can keep this up or it dies like my other projects. chop chop !!!