BimDown for Openclaw

An AI-native bridge for reading and creating Building Information Models (BIM) through code-like CSV and SVG files.

novashang
v1.4.1
Apr 10, 2026
0
874
0

Install & Download

1. ClawHub CLI

The fastest way to install a skill directly from the registry.

npx clawhub@latest install bimdown-cli

2. Manual Installation

Copy the skill folder to one of these locations

Global
~/.openclaw/skills/
Workspace
<project>/skills/

Priority: Workspace > Local > Bundled

3. Prompt Installation

Copy this prompt to OpenClaw to install it automatically.

Help me install bimdown-cli using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).

Prefer to download?

Get the raw skill files in a ZIP archive.

What is BimDown?

BimDown is a specialized bridge between AI agents and building data, designed to make architectural design as intuitive as writing code. By utilizing a dual-file architecture, it stores semantic properties in CSV files and 2D geometry in sibling SVG files. This approach allows developers to model houses, perform quantity surveying, and conduct model reviews using standard developer tools. As part of the Openclaw Skills ecosystem, it enables AI agents to handle real-world dimensions and spatial relationships with high precision.

The skill operates on a METERS-first coordinate system and automates complex architectural calculations. It provides a structured environment where spatial layouts are reasoned through code-like definitions, making it an essential tool for automated construction documentation and architectural programming within Openclaw Skills.

BimDown Use Cases

  • Executing full architectural designs from a text-based brief or requirements program.
  • Modeling existing buildings from floor plan images, sketches, or known dimensions.
  • Performing automated quantity surveying and model audits using SQL-based queries.
  • Generating 2D blueprint renders and 3D shareable previews for design verification.

How BimDown Works

  1. The agent plans the spatial layout, identifying wall positions, room adjacencies, and element IDs.
  2. 2D geometry is written to SVG files using Cartesian coordinates within a standardized Y-axis flip group.
  3. Semantic attributes such as materials and thickness are defined in corresponding CSV files, linked by unique element IDs.
  4. The build command is executed to validate the schema, check wall connectivity, and automatically compute space boundaries.
  5. Visual verification is performed by rendering the project into a PNG blueprint to confirm the layout.
  6. The final project can be published to a secure backend to generate a shareable 3D preview URL for stakeholders.

BimDown Setup

This skill requires the bimdown binary from the bimdown-cli npm package. Verify your environment or install it using the following commands:

# Check for existing installation
which bimdown

# Install if missing (requires user permission)
npm install -g bimdown-cli

Ensure a project_metadata.json file is present in your project root to initialize the environment correctly for Openclaw Skills usage.

BimDown Data Schema & Taxonomy

BimDown organizes data through a strict directory and file taxonomy to ensure compatibility across Openclaw Skills workflows:

Directory File Types Responsibility
global/ .csv Global definitions like level.csv, grid.csv, and mesh.csv.
lv-N/ .csv + .svg Per-level elements like walls and slabs; SVG handles geometry while CSV handles properties.
root project_metadata.json Project configuration, units (meters), and versioning.

Note: Computed fields such as length, area, volume, and level_id are automatically handled by the CLI and should never be manually written to the CSV files.

BimDown Advanced Features

  • SQL Querying: Use DuckDB-powered SQL to query any element table, including SVG-derived virtual columns like length and area.
  • Topology Resolution: Automatically detect coincident endpoints for MEP curves to generate nodes and fill connectivity fields.
  • Structural Differing: Compare two project snapshots to emit structural differences using the diff command.
  • Anonymous Publishing: Share projects with a single command to generate 3D previews without requiring an account or API token.
  • Multi-Project Merging: Combine multiple project directories into a single model while automatically resolving potential ID conflicts.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*