Skip to main content
skill CX Codex

spatial — Claude Code skill: what it does, how to install

Answers spatial data questions using DuckDB and Overture Maps. It handles distance math, polygon lookups, and geospatial file conversions.

Install into Claude Code · Skills CLI (Vercel) All methods and platforms ↓
npx skills add duckdb/duckdb-skills --skill spatial -a claude-code
Open on GitHub

What spatial does

This skill runs DuckDB spatial queries through Bash. When you ask about real-world points of interest, roads, or boundaries without supplying a file, it pulls data directly from Overture Maps on S3. No API key is needed for Overture.

Inside, the skill relies on DuckDB's spatial extension, httpfs, and optional community H3 hex binning. It applies bounding-box filters on Parquet columns before running spatial functions. This pushdown avoids pulling entire datasets across the network.

Coordinate ordering matters here. Setting geometry_always_xy = true keeps longitude and latitude in order, and distance checks require POINT_2D on the WGS84 ellipsoid. A local DuckDB installation is required; if missing, the CLI must be installed first.

What it helps with

  • calculate real-world ellipsoid distances with ST_Distance_Spheroid
  • query POIs, roads, or buildings from Overture Maps on S3
  • check point-in-polygon containment across spatial files
  • convert datasets between GeoJSON, Shapefile, and GeoPackage

Example prompts for the agent

  • Find all pharmacies within this bounding box using Overture Maps
  • Join points from customers.csv with polygons in neighborhoods.geojson
  • Convert track.gpx into a GeoPackage file using DuckDB

How to install spatial

1Skills CLI (Vercel)

npx skills add duckdb/duckdb-skills --skill spatial -a claude-code

2Manually: put the files in place

~/.claude/skills/spatial/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/spatial/

Step by step, with commands for each platform: How to install skills in Claude Code: five ways

Questions about spatial

How to install spatial in Claude Code?

Run npx skills add duckdb/duckdb-skills --skill spatial -a claude-code.

What does the spatial skill do?

It executes DuckDB SQL queries to analyze coordinates, convert geospatial files, and query global Overture Maps data on S3.

Why is geometry_always_xy enabled by default?

It guarantees that functions read coordinates as longitude followed by latitude. Without this setting, spheroid functions assume latitude first and produce incorrect spatial calculations.

Source text

SKILL.md · MIT
---
name: spatial
description: >
  Answer questions about spatial data using DuckDB. Use when the user mentions locations,
  coordinates, lat/lng, distances, maps, addresses, "near", "within", "closest", geographic
  names, or spatial file formats (GeoJSON, Shapefile, GeoPackage, GPX, GeoParquet). Also
  triggers when the user wants to find places, buildings, or roads — Overture Maps provides
  free global data on S3 with zero API keys. Handles spatial joins, distance calculations,
  containment checks, density analysis, and format conversions for geographic data.
argument-hint: <question or file> [additional context]
allowed-tools: Bash
---

You are answering spatial questions using DuckDB's spatial extension and, when needed, Overture Maps as a free global data source.

Question or file: `$0`
Additional context: `${1:-}`

## Step 1 — Understand what the user needs

Classify the question:

| Pattern | Data source | Key functions |
|---------|-------------|---------------|
| "Find X near Y" (no user file) | Overture Maps on S3 | `ST_Distance_Spheroid`, bbox filtering |
| "How far between A and B" | Geocode or user data | `ST_Distance_Spheroid` |
| "Which points fall inside polygons" | User files | `ST_Contains` |
| "Analyze this GeoJSON/Shapefile/GPX" | User file | `ST_Read`, measurement functions |
| "Show density/hotspots" | User or Overture data | H3 hex binning |
| "Convert to GeoJSON/GeoPackage" | User file | `COPY TO (FORMAT GDAL)` |
| "Count buildings/roads in area" | Overture Maps | bbox filtering + aggregation |

If the question involves real-world places, POIs, buildings, roads, or boundaries and the user hasn't provided a file, use **Overture Maps** — read `references/overture.md` for S3 paths and schema.

For spatial function syntax, read `references/functions.md`.

## Step 2 — Write and run the query

Always start with:
```sql
LOAD spatial;
SET geometry_always_xy = true;
```

Add extensions as needed:
- Overture/remote data: `LOAD httpfs; CREATE SECRET (TYPE S3, PROVIDER config, REGION 'us-west-2');`
- H3 hex binning: `INSTALL h3 FROM community; LOAD h3;`

### Key principles

**bbox filtering first** — When querying Overture, always filter on `bbox.xmin/xmax/ymin/ymax` before any spatial function. This uses Parquet predicate pushdown and avoids downloading the full dataset.

**Always set `geometry_always_xy = true`** — This ensures all spatial functions interpret coordinates as longitude, latitude (the standard for Overture, GeoJSON, and most data sources). Without it, spheroid functions assume latitude first and return wrong results.

**Use spheroid functions for real-world distances** — `ST_Distance_Spheroid` returns meters on the WGS84 ellipsoid. Plain `ST_Distance` uses planar coordinates and gives meaningless results for lat/lng. **Important:** spheroid functions (`ST_Distance_Spheroid`, `ST_Area_Spheroid`, etc.) require `POINT_2D` inputs, not generic `GEOMETRY`. Overture geometry columns are typed `GEOMETRY('OGC:CRS84')` and cannot be cast directly. Extract coordinates first:
```sql
ST_Point(ST_X(geometry), ST_Y(geometry))::POINT_2D
```

**CSV with lat/lng needs conversion** — `ST_Point(longitude, latitude)` (longitude first). This is the most common gotcha.

Run the query in a single bash call:

```bash
duckdb -c "
LOAD spatial;
<ADDITIONAL_SETUP>
<YOUR_QUERY>
"
```

## Step 3 — Present results

- For tabular results: show the data directly
- For spatial results: consider exporting to GeoJSON for visualization (`COPY TO 'result.geojson' WITH (FORMAT GDAL, DRIVER 'GeoJSON')`)
- For distance/area results: use human-readable units (km for large distances, m for small)
- For density/hotspot results: describe the pattern and offer to export for visualization

If the query fails:
- **`duckdb: command not found`** → delegate to `/duckdb-skills:install-duckdb`
- **Missing extension** → `INSTALL spatial; LOAD spatial;` or `INSTALL h3 FROM community; LOAD h3;`
- **S3 access denied** → suggest checking AWS credentials
- **No results with Overture** → widen the bbox, check the category spelling, or try a broader search

© duckdb · MIT · duckdb/duckdb-skills

Similar skills

All skills for Data and databases →

From the same set7

Project sponsors

Companies that keep this analytics open
Gold sponsor

Impressions, clicks and code copies — open statistics. We keep the numbers public for the same reason NeBlask itself is public. Become a sponsor