一键导入
process-raster
// Process raster data: clip by bounding box, stack multiple bands, mosaic GeoTIFFs, or convert between raster and vector formats.
// Process raster data: clip by bounding box, stack multiple bands, mosaic GeoTIFFs, or convert between raster and vector formats.
Run pre-trained AI models on geospatial imagery. Detect buildings, cars, ships, solar panels, agriculture fields, or use text-prompted segmentation with GroundedSAM. Requires GPU for best performance.
Download NAIP aerial imagery for a bounding box. Specify coordinates as minx,miny,maxx,maxy in WGS84 and optionally a year.
Inspect any raster or vector geospatial file. Returns CRS, bounds, bands, resolution, dtype, attribute summaries, and band statistics. Supports GeoTIFF, Shapefile, GeoJSON, GeoPackage, GeoParquet, and more.
Verify that the geoai Python package is installed and functional. If not, provide installation instructions. Optionally check extra dependencies for deep learning models.
Download Overture Maps data (buildings, places, roads, land use, water, etc.) for a bounding box. Returns a GeoDataFrame saved as GeoJSON or GeoPackage.
Search past Claude Code session logs to recover context from previous conversations. Finds past decisions, data paths, CRS info, model configurations, and unresolved work. Works across all projects or scoped to the current one.
| name | process-raster |
| description | Process raster data: clip by bounding box, stack multiple bands, mosaic GeoTIFFs, or convert between raster and vector formats. |
| argument-hint | <operation> <input> [options] |
| allowed-tools | Bash |
You are helping the user process geospatial raster data using geoai.
Input: $@
Follow these steps in order.
Parse $@ to identify the requested operation:
| Operation | Triggers | Required inputs |
|---|---|---|
clip | "clip", "crop", "subset", --bbox present | input raster + bbox |
stack | "stack", "combine bands" | list of input rasters |
mosaic | "mosaic", "merge" | input directory or list of rasters |
raster-to-vector | "to vector", "vectorize", "polygonize" | input raster |
vector-to-raster | "to raster", "rasterize", "burn" | input vector + pixel size |
If the operation is unclear from the input, ask the user to specify.
For single-file operations (clip, raster-to-vector, vector-to-raster):
find "$PWD" -name "INPUT_FILENAME" -not -path '*/.git/*' 2>/dev/null
For multi-file operations (stack, mosaic), if a directory is given:
find "INPUT_DIR" -name "*.tif" -o -name "*.tiff" 2>/dev/null | sort
If the user recently inspected or downloaded a file and did not specify an input, check the state file for context:
STATE_DIR=""
test -f .geoai-skills/state.json && STATE_DIR=".geoai-skills"
PROJECT_ROOT="$(git rev-parse --show-toplevel 2>/dev/null || echo "$PWD")"
PROJECT_ID="$(echo "$PROJECT_ROOT" | tr '/' '-')"
test -f "$HOME/.geoai-skills/$PROJECT_ID/state.json" && STATE_DIR="$HOME/.geoai-skills/$PROJECT_ID"
If state exists, read the last inspected or downloaded file:
python3 -c "
import json
with open('STATE_DIR/state.json') as f:
state = json.load(f)
if 'last_inspected' in state:
print(f'Last inspected: {state[\"last_inspected\"][\"path\"]}')
if 'downloaded_files' in state:
for f in state['downloaded_files']:
print(f'Downloaded: {f}')
"
python3 -c "
import geoai
result = geoai.clip_raster_by_bbox(
input_raster='INPUT_PATH',
output_raster='OUTPUT_PATH',
bbox=[MINX, MINY, MAXX, MAXY],
)
print(f'Clipped raster saved to: {result}')
info = geoai.get_raster_info(result)
for k, v in info.items():
print(f'{k}: {v}')
"
Default output: ./clipped_<original_name>.tif
python3 -c "
import geoai
result = geoai.stack_bands(
input_files=['FILE1', 'FILE2', 'FILE3'],
output_file='OUTPUT_PATH',
)
print(f'Stacked raster saved to: {result}')
info = geoai.get_raster_info(result)
for k, v in info.items():
print(f'{k}: {v}')
"
python3 -c "
import geoai
result = geoai.mosaic_geotiffs(
input_dir='INPUT_DIR',
output_file='OUTPUT_PATH',
)
print(f'Mosaic saved to: {result}')
info = geoai.get_raster_info(result)
for k, v in info.items():
print(f'{k}: {v}')
"
python3 -c "
import geoai
gdf = geoai.raster_to_vector(
raster_path='INPUT_PATH',
output_path='OUTPUT_PATH',
)
print(f'Vectorized: {len(gdf)} features')
print(f'Saved to: OUTPUT_PATH')
print(f'Columns: {list(gdf.columns)}')
"
Default output: ./<original_name>.gpkg
python3 -c "
import geoai
result = geoai.vector_to_raster(
vector_path='INPUT_PATH',
output_path='OUTPUT_PATH',
pixel_size=PIXEL_SIZE,
)
print(f'Rasterized: {result}')
info = geoai.get_raster_info(result)
for k, v in info.items():
print(f'{k}: {v}')
"
Default pixel size: 1.0 (or infer from context). Default output: ./<original_name>.tif
Replace all placeholder values with actual paths and parameters before running.
If a state directory exists, update it with the output file path using the same state resolution pattern as Step 2.
Report:
Then suggest: "Use /geoai-skills:inspect-geo to examine the result in detail."
import geoai fails -> delegate to /geoai-skills:install-geoai.find to locate, suggest corrected path.