Metadata Extraction

Pull metadata from translated Revit🏠RevitAutodesk's BIM software for architecture and construction.View in glossary models for analysis or reporting.

Workflow Overview

CLI Approach

Step 1: Get Model URN

URN=$(raps object urn my-bucket building.rvt --output plain)

Step 2: Check Translation Status

raps translate manifest "$URN"

Step 3: Extract All Metadata

# Model metadata (GUID, views, etc.)
raps derivative metadata "$URN" --output json > metadata.json

# Model hierarchy
raps derivative tree "$URN" --output json > model-tree.json

# All properties (BIM data)
raps derivative properties "$URN" --output json > properties.json

# Available views
raps derivative views "$URN" --output json > views.json

Step 4: Analyze Data

# Count elements by category
cat properties.json | jq 'group_by(.category) | .[] | {category: .[0].category, count: length}'

# Export specific properties to CSV
cat properties.json | jq -r '
  ["ObjectId", "Name", "Category"],
  (.[] | [.objectid, .name, .category])
  | @csv' > elements.csv

CI/CD Pipeline

# .github/workflows/metadata-extraction.yml
name: BIM Metadata Extraction

on:
  workflow_dispatch:
    inputs:
      bucket:
        description: 'Bucket name'
        required: true
      model:
        description: 'Model filename'
        required: true

jobs:
  extract-metadata:
    runs-on: ubuntu-latest
    steps:
      - name: Install RAPS
        run: cargo install raps

      - name: Extract all metadata
        env:
          APS_CLIENT_ID: ${{ secrets.APS_CLIENT_ID }}
          APS_CLIENT_SECRET: ${{ secrets.APS_CLIENT_SECRET }}
        run: |
          URN=$(raps object urn "${{ inputs.bucket }}" "${{ inputs.model }}" --output plain)

          mkdir -p ./metadata-export

          raps derivative metadata "$URN" --output json > ./metadata-export/metadata.json
          raps derivative tree "$URN" --output json > ./metadata-export/tree.json
          raps derivative properties "$URN" --output json > ./metadata-export/properties.json
          raps derivative views "$URN" --output json > ./metadata-export/views.json

      - name: Generate summary
        run: |
          echo "# Metadata Summary" > ./metadata-export/summary.md
          echo "Model: ${{ inputs.model }}" >> ./metadata-export/summary.md
          echo "Elements: $(jq 'length' ./metadata-export/properties.json)" >> ./metadata-export/summary.md
          echo "Views: $(jq 'length' ./metadata-export/views.json)" >> ./metadata-export/summary.md

      - name: Upload artifacts
        uses: actions/upload-artifact@v4
        with:
          name: bim-metadata-${{ github.run_id }}
          path: ./metadata-export/

Pipeline Flow

MCP Integration

Extract and Summarize

User: "Extract all metadata from building.rvt and summarize the room count"

AI Response:

I found 47 rooms in the model:

• Level 1: 15 rooms (3,450 sq ft total)
• Level 2: 18 rooms (4,200 sq ft total)
• Level 3: 14 rooms (3,100 sq ft total)

Compare Models

User: "Compare the element counts between v1 and v2 of the building model"

AI Response:

Model Comparison:

Category	v1	v2	Change
Walls	245	258	+13
Doors	89	92	+3
Windows	156	156	0
Rooms	47	52	+5

Common Queries

Count by Category

cat properties.json | jq '
  group_by(.category) |
  map({category: .[0].category, count: length}) |
  sort_by(-.count)'

Find Specific Elements

# Find all doors
cat properties.json | jq '[.[] | select(.category == "Doors")]'

# Find elements by name pattern
cat properties.json | jq '[.[] | select(.name | test("Wall.*Exterior"))]'

Export to Database Format

# Generate SQL inserts
cat properties.json | jq -r '.[] |
  "INSERT INTO elements (id, name, category) VALUES (\(.objectid), '\''\(.name)'\'', '\''\(.category)'\'');"'

Related

• Room & Space Export
• Model Coordination Pipeline
• Cookbook: AEC & BIM