Build the application

## Build
Analyze wildland-urban interface commercial property perimeters by locating recent high-resolution earth observation imagery via STAC metadata. Generate vegetation trimming work orders to ensure defensible space compliance.

## User
Facility Risk Specialist

## Workflow
1. Extract property boundary vectors and defined 30-meter and 100-meter defensible perimeter buffers.
2. Query a STAC API for Items covering the property boundaries and extract image asset download URLs.
3. Calculate canopy density anomalies and dry fuel accumulation within defensible zones using the retrieved images.
4. Generate an actionable groundskeeping work order specifying clearing quadrants and debris disposal methods.

## Interface
Facility management portal visualizing vegetation encroachment overlays alongside clear contractor task lists.

## Data
Utilize STAC API compliant endpoints (e.g. Earth Search) to query metadata and download associated image assets for pre- and post-season analysis.
Starter examples below are synthetic. They are not downloaded records.

## Source
[STAC specification](https://raw.githubusercontent.com/radiantearth/stac-spec/master/README.md)
Parses STAC Item metadata, specifically bounding boxes, datetime fields, and asset download URLs conforming to STAC specification 1.1.0.

## Tools
qdrant, pystac, shapely.
Indexes visual and spatial parcel canopy embedding vectors to retrieve and cluster high-risk vegetation encroachment patterns.
Local or cloud multimodal model capable of interpreting satellite vegetation metadata and writing clean operational checklists.

## Deliverables
Submit runnable source, pinned dependencies, setup commands, a working interface, example outputs, and a recorded end-to-end demo. Include tests and an action log.

## Acceptance
- Detects vegetation canopy coverage exceeding 20% in the immediate 30-meter structural defense zone.
- Emits localized clearing work tickets mapped to parcel perimeter cardinal zones (North, South, East, West).
- Includes contractor safety protocols regarding deadwood removal without authorizing unsupervised tree felling.

## Starter Examples
```json
[
  {
    "input": {
      "property_id": "FAC-WUI-501",
      "zone_30m_canopy_pct": 34.5,
      "zone_100m_canopy_pct": 52,
      "predominant_fuel": "chaparral_scrub",
      "quadrant_flags": {
        "north": true,
        "south": false,
        "east": true,
        "west": false
      }
    },
    "expected_output": {
      "property_id": "FAC-WUI-501",
      "compliance_status": "non_compliant",
      "work_orders": [
        {
          "quadrant": "north",
          "target_reduction_pct": 14.5,
          "clearance_type": "canopy_thinning_and_brushing"
        },
        {
          "quadrant": "east",
          "target_reduction_pct": 14.5,
          "clearance_type": "deadwood_and_brush_removal"
        }
      ],
      "human_inspection_required": true
    }
  },
  {
    "input": {
      "property_id": "FAC-WUI-502",
      "zone_30m_canopy_pct": 8,
      "zone_100m_canopy_pct": 22,
      "predominant_fuel": "mowed_turf",
      "quadrant_flags": {
        "north": false,
        "south": false,
        "east": false,
        "west": false
      }
    },
    "expected_output": {
      "property_id": "FAC-WUI-502",
      "compliance_status": "compliant",
      "work_orders": [],
      "human_inspection_required": false
    }
  }
]
```

## Safety
Outputs vegetation management recommendations only; does not issue binding regulatory citations or unsupervised forestry permits.
Use mock external writes. Require human review before any real deployment.