Claude Streetview — Data skill for Claude Code

  ___ _                _   __ _
 / __| |_ _ _ ___ ___| |_ \ \  ()___ __ __
 \__ \  _| '_/ -_) -_)  _| \ \ | / -_) V  V
 |___/\__|_| \___\___|\__| \_\_|_\___|\_/\_/

 street-level imagery + 3D reference for Claude

A [Claude Skill](https://docs.claude.com) for pulling street-level photos and street-context data (geometry, buildings, street furniture, elevation) from free and paid imagery providers, and packaging it for 3D reconstruction work.

Why

Asking an LLM to "help model this street" usually means it has to guess. This gives it real data instead: georeferenced photos, camera headings, building footprints, and OSM street furniture, bundled into one folder a 3D tool can build against.

Providers

Provider	Key required	Notes
KartaView	no	default, always available
Panoramax	no	default, always available
Mapillary	free token	better coverage, 360s, object detections
Google Street View	paid API key	best global coverage, restrictive license

If no key is configured, only the two free-and-open providers are used. Add a key and it's picked up automatically — nothing else changes.

Install

pip install -r requirements.txt

Or, as an installable command:

pip install .
streetview --help

Usage

# what's currently configured and reachable
python scripts/streetview.py config-doctor

# photos near a point, across every available provider
python scripts/streetview.py search 48.8566,2.3522

# every photo around one point, sorted by compass heading (a 360° sweep)
python scripts/streetview.py orbit 48.8566,2.3522

# bundle photos + camera poses + street geometry + buildings + street
# furniture + elevation into one folder, for 3D modelling reference
python scripts/streetview.py --radius 80 reference-pack 48.8566,2.3522 ./out/my_street

Optional API keys:

python scripts/streetview.py config-set mapillary_token 
python scripts/streetview.py config-set google_