Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Convert RVT/RFA files to Excel databases. Extract BIM element data, properties, and quantities.
.claude/skills/datadrivenconstruction-rvt-to-excel/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 2% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 104% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 54% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 26% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 36% | 0% |
BIM data inside RVT files needs to be extracted for:
Convert RVT files to structured Excel databases for analysis and reporting.
bashRvtExporter.exe <input_path> [export_mode] [options]
| Mode | Categories | Description | |------|-----------|-------------| | basic | 309 | Essential structural elements | | standard | 724 | Standard BIM categories | | complete | 1209 | All Revit categories | | custom | User-defined | Specific categories only |
| Option | Description | |--------|-------------| | bbox | Include bounding box coordinates | | rooms | Include room associations | | schedules | Export all schedules to sheets | | sheets | Export sheets to PDF |
bash# Basic export RvtExporter.exe "C:\Projects\Building.rvt" basic # Complete with bounding boxes RvtExporter.exe "C:\Projects\Building.rvt" complete bbox # Full export with all options RvtExporter.exe "C:\Projects\Building.rvt" complete bbox rooms schedules sheets # Batch processing for /R "C:\Projects" %f in (*.rvt) do RvtExporter.exe "%f" standard bbox
pythonimport subprocess import pandas as pd from pathlib import Path from typing import List, Optional class RevitExporter: def __init__(self, exporter_path: str = "RvtExporter.exe"): self.exporter = Path(exporter_path) if not self.exporter.exists(): raise FileNotFoundError(f"RvtExporter not found: {exporter_path}") def convert(self, rvt_file: str, mode: str = "complete", options: List[str] = None) -> Path: """Convert Revit file to Excel.""" rvt_path = Path(rvt_file) if not rvt_path.exists(): raise FileNotFoundError(f"Revit file not found: {rvt_file}") cmd = [str(self.exporter), str(rvt_path), mode] if options: cmd.extend(options) result = subprocess.run(cmd, capture_output=True, text=True) if result.returncode != 0: raise RuntimeError(f"Export failed: {result.stderr}") # Output file is same name with .xlsx extension output_file = rvt_path.with_suffix('.xlsx') return output_file def batch_convert(self, folder: str, mode: str = "standard", pattern: str = "*.rvt") -> List[Path]: """Convert all Revit files in folder.""" folder_path = Path(folder) converted = [] for rvt_file in folder_path.glob(pattern): try: output = self.convert(str(rvt_file), mode) converted.append(output) print(f"Converted: {rvt_file.name}") except Exception as e: print(f"Failed: {rvt_file.name} - {e}") return converted def read_elements(self, xlsx_file: str) -> pd.DataFrame: """Read converted Excel as DataFrame.""" return pd.read_excel(xlsx_file, sheet_name="Elements") def get_quantities(self, xlsx_file: str, group_by: str = "Category") -> pd.DataFrame: """Get quantity summary grouped by category.""" df = self.read_elements(xlsx_file) # Group and count summary = df.groupby(group_by).agg({ 'ElementId': 'count', 'Area': 'sum', 'Volume': 'sum' }).reset_index() summary.columns = [group_by, 'Count', 'Total_Area', 'Total_Volume'] return summary
| Sheet | Content | |-------|---------| | Elements | All BIM elements with properties | | Categories | Element categories summary | | Levels | Building levels | | Materials | Material definitions | | Parameters | Shared parameters |
| Column | Type | Description | |--------|------|-------------| | ElementId | int | Unique Revit ID | | Category | string | Element category | | Family | string | Family name | | Type | string | Type name | | Level | string | Associated level | | Area | float | Surface area (m²) | | Volume | float | Volume (m³) | | BBox_MinX/Y/Z | float | Bounding box min | | BBox_MaxX/Y/Z | float | Bounding box max |
python# Initialize exporter exporter = RevitExporter("C:/Tools/RvtExporter.exe") # Convert single file xlsx = exporter.convert("C:/Projects/Office.rvt", "complete", ["bbox", "rooms"]) # Read and analyze df = exporter.read_elements(str(xlsx)) print(f"Total elements: {len(df)}") # Quantity summary quantities = exporter.get_quantities(str(xlsx)) print(quantities) # Export to CSV for further processing df.to_csv("elements.csv", index=False)
python# Full pipeline: Revit → Excel → Cost Estimate from semantic_search import CWICRSemanticSearch # 1. Convert Revit exporter = RevitExporter() xlsx = exporter.convert("project.rvt", "complete", ["bbox"]) # 2. Extract quantities df = exporter.read_elements(str(xlsx)) quantities = df.groupby('Category')['Volume'].sum().to_dict() # 3. Search CWICR for pricing search = CWICRSemanticSearch() costs = {} for category, volume in quantities.items(): results = search.search_work_items(category, limit=5) if not results.empty: avg_price = results['unit_price'].mean() costs[category] = volume * avg_price print(f"Total estimate: ${sum(costs.values()):,.2f}")
basic for quick analysis, complete for full data| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 14,558 | 3,449 | -76% | 1 | 1 | 0% | 2,339 | 2,393 | +2% | 0 | 0 | — |
case-02 | fail→pass | 46,307 | 11,041 | -76% | 1 | 1 | 0% | 1,987 | 4,053 | +104% | 0 | 0 | — |
case-03 | fail→pass | 10,955 | 7,770 | -29% | 1 | 1 | 0% | 2,093 | 3,231 | +54% | 0 | 0 | — |
case-04 | fail→pass | 11,820 | 3,074 | -74% | 1 | 1 | 0% | 1,758 | 2,210 | +26% | 0 | 0 | — |
case-05 | fail→pass | 31,359 | 2,529 | -92% | 1 | 1 | 0% | 1,548 | 2,106 | +36% | 0 | 0 | — |
case-06 | fail→pass | 13,962 | 5,168 | -63% | 1 | 1 | 0% | 2,488 | 2,598 | +4% | 0 | 0 | — |
case-07 | pass→pass | 10,966 | 4,274 | -61% | 1 | 1 | 0% | 1,968 | 2,372 | +21% | 0 | 0 | — |
case-08 | fail→pass | 16,256 | 2,907 | -82% | 1 | 1 | 0% | 2,577 | 2,167 | -16% | 0 | 0 | — |
case-09 | fail→pass | 16,027 | 3,923 | -76% | 1 | 1 | 0% | 2,529 | 2,449 | -3% | 0 | 0 | — |
case-10 | pass→pass | 12,355 | 2,050 | -83% | 1 | 1 | 0% | 1,980 | 2,070 | +5% | 0 | 0 | — |
case-11 | fail→pass | 8,720 | 2,460 | -72% | 1 | 1 | 0% | 1,459 | 2,152 | +47% | 0 | 0 | — |
case-12 | fail→pass | 13,863 | 1,894 | -86% | 1 | 1 | 0% | 2,055 | 1,998 | -3% | 0 | 0 | — |
case-13 | pass→pass | 13,049 | 2,002 | -85% | 1 | 1 | 0% | 1,997 | 1,960 | -2% | 0 | 0 | — |
case-14 | fail→pass | 12,538 | 1,598 | -87% | 1 | 1 | 0% | 1,808 | 1,916 | +6% | 0 | 0 | — |
case-15 | fail→pass | 9,239 | 2,010 | -78% | 1 | 1 | 0% | 1,448 | 1,976 | +36% | 0 | 0 | — |
case-16 | fail→pass | 11,180 | 8,560 | -23% | 1 | 1 | 0% | 2,046 | 3,408 | +67% | 0 | 0 | — |
case-17 | pass→pass | 22,038 | 13,472 | -39% | 1 | 1 | 0% | 4,214 | 4,315 | +2% | 0 | 0 | — |
case-18 | pass→pass | 14,686 | 3,318 | -77% | 1 | 1 | 0% | 2,165 | 2,299 | +6% | 0 | 0 | — |
case-19 | fail→fail | 16,115 | 9,004 | -44% | 1 | 1 | 0% | 2,535 | 3,411 | +35% | 0 | 0 | — |
case-25 | pass→pass | 20,133 | 18,882 | -6% | 1 | 1 | 0% | 3,344 | 4,937 | +48% | 0 | 0 | — |
case-20 | pass→pass | 8,069 | 4,055 | -50% | 1 | 1 | 0% | 1,317 | 2,419 | +84% | 0 | 0 | — |
case-21 | pass→pass | 13,750 | 1,916 | -86% | 1 | 1 | 0% | 1,986 | 2,026 | +2% | 0 | 0 | — |
case-22 | pass→pass | 14,881 | 15,087 | +1% | 1 | 1 | 0% | 2,372 | 4,408 | +86% | 0 | 0 | — |
case-23 | pass→pass | 13,509 | 13,058 | -3% | 1 | 1 | 0% | 2,432 | 4,162 | +71% | 0 | 0 | — |
case-24 | pass→pass | 12,976 | 11,037 | -15% | 1 | 1 | 0% | 2,324 | 3,868 | +66% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 25 cases were attempted, and 23 counted toward the lift figure. The other 2 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +52 percentage points is the difference between those two pass rates over the 23 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.