AN AGENT SKILL FOR POLISHED MAP DELIVERY

Spatial data in. Map product out.

Turn Excel, CSV, GeoJSON, GeoPackage, and Shapefiles into searchable interactive maps, slide-ready images, and publication figures—with Atlas Studio Light visual defaults, inspection, provenance, and browser verification built in.

$ interactive-map-builder inspect data.geojson
✓ CRS · fields · geometry
$ interactive-map-builder build map_spec.json
✓ map.html · PNG · SVG · PDF
$ interactive-map-builder verify --dist dist
✓ browser-ready deliverable
REAL DATA · REAL GENERATED HTML

Two map products, not two screenshots.

Both demos are built from fixed NYC Open Data snapshots by the deterministic engine shipped in this repository.

MAP + LIST

Lower Manhattan parcel explorer

Search addresses, filter land-use classes and numeric fields, sort records, and inspect a selected parcel.

Open live map ↗
MULTILAYER

Downtown Brooklyn mobility context

Inspect subway stations, bicycle routes, and neighborhoods with independent visibility and layer-specific search.

Open live map ↗
ONE AUDITABLE DELIVERY PATH

Inspect, confirm, specify, build, and verify.

The Skill resolves ambiguity, Atlas Studio Light supplies a coordinated starting point, and the Python engine owns deterministic data handling and rendering.

01

Inspect

Discover layers, geometry, CRS, field roles, and blocking issues.

02

Confirm

Close only the information gaps the data cannot answer.

03

Resolve

Apply geometry-aware, density-aware visual defaults while preserving explicit choices.

04

Build

Create a single-file Leaflet app and static figures.

05

Verify

Exercise browser behavior, counts, hashes, and provenance.