# Dan Weist — Load Boundaries Into Field Audit Tool
**Agent:** engineer
**Priority:** high
## Goal
Take all of Dan Weist's boundary shapefiles and load them into the field naming audit tool so Doug can review them at https://farmiq.ai/field-audit/
## Step 1: Create customer directory
mkdir -p /data/Sandbox/farmiq.ai/field-audit/data/customers/dan/sources/gdrive
## Step 2: Collect all boundary shapefiles
Query the GDrive manifest for Dan's processed boundaries:
sqlite3 /data/agentpi/state/gdrive_scan.db "SELECT absolute_path FROM gdrive_files WHERE customer='weist,_dan' AND file_type='boundary_shapefile' AND status='ok'"
For each .shp file, convert to GeoJSON (using ogr2ogr or Python with fiona/shapely) and save to:
/data/Sandbox/farmiq.ai/field-audit/data/customers/dan/sources/gdrive/{field_name}.geojson
Try to extract a meaningful field name from the file path (e.g., "w01_Boundary.shp" → "W01").
Also collect any planted_boundary.geojson files from /data/processed_data/weist,_dan/ and copy those to:
/data/Sandbox/farmiq.ai/field-audit/data/customers/dan/sources/planted/
## Step 3: Run the audit engine
python /data/Sandbox/FieldNamingAudit/field_naming_audit.py --scan /data/Sandbox/farmiq.ai/field-audit/data/customers/dan/sources --output /data/Sandbox/farmiq.ai/field-audit/data/customers/dan/audit_results
## Step 4: Report
Post to dashboard how many boundaries were loaded and if the audit ran successfully.
Tell Doug the audit is ready to view at the field-audit URL.
## Notes
- There are 1,759 boundary shapefiles — many will be duplicates or coverage maps, not true field boundaries
- Files named *_Boundary.shp are actual field boundaries
- Files named *_Coverage.shp are spray/seeding coverage — still useful but different
- Data goes back to 2004 — expect old formats and projections
- The audit tool will handle dedup and matching