ifcopenshell-mcp

ifcopenshell-mcp

An MCP server that lets users query, validate, convert, and compare IFC/BIM files using natural language. Runs fully locally and integrates with AI agents like Claude.

Category
Visit Server

README

ifcopenshell-mcp

License: MIT Python 3.10+ MCP Compatible

Talk to your BIM models in natural language. An open-source MCP server for IfcOpenShell — query, validate, convert, and compare IFC files through AI agents.


For Architects (no coding required)

What does this do?

Drag an .ifc file into Claude and ask questions like:

"How many fire doors are on Level 2?"

Claude reads your model and answers:

Found 2 IfcDoor elements with FireRating=R60:
- Fire Exit GF (Ground Floor)
- Fire Exit L2 (Level 2)

No Revit. No Navisworks. No SQL. Just drag, ask, get answers.

What can I ask?

You ask... It does...
"Summarize this model" Shows project name, total elements, spatial structure
"How many walls on Ground Floor?" Counts and lists them with IDs
"Validate that all elements have materials" QA/QC report: PASS / FAIL / WARNING per element
"Extract concrete quantities for Level 1" Area, volume, length organized by element
"What changed since last week's version?" Diff report: added, removed, modified elements
"Convert this to GLB for web viewing" Creates a 3D file viewable in any browser

Complete tool list at the bottom of this page.

How to install (5 minutes, one time only)

Step 1: Install uv (a tiny installer). Open your terminal once and paste:

Windows (PowerShell):

powershell -c "irm https://astral.sh/uv/install.ps1 | iex"

Mac (Terminal):

curl -LsSf https://astral.sh/uv/install.sh | sh

Step 2: Download this tool. In the same terminal:

git clone https://github.com/Daviidro/ifcopenshell-mcp.git
cd ifcopenshell-mcp
uv sync

Step 3: Tell Claude about it. Open Claude Desktop → Settings → Developer → Edit Config, and paste:

{
  "mcpServers": {
    "ifcopenshell": {
      "command": "uv",
      "args": ["run", "ifcopenshell-mcp"],
      "cwd": "D:/ifcopenshell-mcp"
    }
  }
}

Restart Claude Desktop. Done. Drag any .ifc file and start asking.

Is this safe?

Yes. Your files never leave your computer. Period.

  • Everything runs 100% locally — no cloud, no accounts, no internet needed
  • It only reads IFC files you explicitly drag into Claude — nothing else
  • It cannot write files outside your project folder
  • It cannot access system files
  • The code is open source (MIT) — anyone can inspect every line
  • Built on IfcOpenShell, trusted by thousands of architects and engineers since 2011

No telemetry. No analytics. No hidden anything.


For Developers (contributing code)

Tools

Tool Description Example question
query_ifc_elements Search by type, level, or properties "Show all fire doors"
get_ifc_statistics Model overview and class distribution "Give me a summary"
get_element_properties Full properties of one element by GUID "Details of wall 3GQqVv..."
list_ifc_spaces All rooms/spaces "List all rooms"
validate_ifc_model QA/QC: materials, properties, containment, naming "Validate this model"
compare_ifc_models Diff between two IFC versions "What changed?"
extract_ifc_quantities Quantity takeoffs (markdown/CSV/JSON) "Concrete volume?"
convert_ifc_to_gltf IFC to GLB for web visualization "Make it viewable online"

Supported formats

  • IFC2x3, IFC4, IFC4.3 (.ifc, .ifczip, .ifcxml)
  • Output: markdown, CSV, JSON
  • GLB: requires IFC files with geometric representations. Semantic-only IFCs report conversion failure.

Setup

git clone https://github.com/Daviidro/ifcopenshell-mcp.git
cd ifcopenshell-mcp
uv sync
uv run pytest tests/ -v    # 19 tests, all must pass

Tests

uv run pytest tests/ -v
# Expected: 19 passed (13 functional + 6 error handling)

Roadmap

  • [ ] Clash detection (bounding box)
  • [ ] COBie spreadsheet export
  • [ ] IFC to Revit roundtrip via Speckle
  • [ ] Classification system validation (Uniclass, OmniClass)

Contributing

See CONTRIBUTING.md. Built on IfcOpenShell (LGPL-3.0) and MCP. Also check out the official IfcOpenShell MCP.

Ecosystem

License

MIT — see LICENSE file.

Recommended Servers

playwright-mcp

playwright-mcp

A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.

Official
Featured
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.

Official
Featured
Local
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.

Official
Featured
Local
TypeScript
VeyraX MCP

VeyraX MCP

Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.

Official
Featured
Local
graphlit-mcp-server

graphlit-mcp-server

The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.

Official
Featured
TypeScript
Kagi MCP Server

Kagi MCP Server

An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.

Official
Featured
Python
Neon Database

Neon Database

MCP server for interacting with Neon Management API and databases

Official
Featured
E2B

E2B

Using MCP to run code via e2b.

Official
Featured
Exa Search

Exa Search

A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.

Official
Featured
Qdrant Server

Qdrant Server

This repository is an example of how to create a MCP server for Qdrant, a vector search engine.

Official
Featured