# ASCII City Engine

## Why Install This Skill

Turn public elevation, building, and street data into a portable, first-person colored-ASCII city without re-deriving coordinate, collision, and provenance rules. The skill keeps the renderer city-neutral and city packs interchangeable.

## What You Get

- A heightfield and solid-footprint engine architecture.
- JSON Schemas for manifests and world tiles.
- A deterministic offline city-pack validator.
- A pure browser, single-file Canvas 2D reference engine.
- A small redistributable downtown Raleigh sample with real public footprints.
- Reproducible GIS ingestion and Raleigh walkthrough references.

## Quick Start

```sh
python3 scripts/validate-city-pack.py assets/raleigh-downtown-sample
python3 -m http.server 8000
```

Then open `http://localhost:8000/assets/ascii-city-engine.html`. The scaffold reads `assets/raleigh-downtown-sample/manifest.json` for the spawn point and first world tile; use W/S or arrow keys to move and A/D or left/right arrows to turn. The sample is dense: named-street signs, traffic signals, crosswalks, trees, and transit stops all render from real recorded data, and the HUD names the street you are standing on and the building you are facing.

## Triggers

Load this skill for requests to build a colored ASCII city renderer, create a terrain-following first-person walker, define a pluggable city pack, ingest public GIS into the provided contract, or reproduce the Raleigh proof of concept.

Do not load it for ordinary GIS analysis, WebGL/3D engines, multi-level interiors or tunnels, gameplay systems, or requests to commit full-resolution GIS archives.

## Requirements

- Python 3 standard library for validation and conversion examples.
- A current desktop browser with Canvas 2D and `fetch` support.
- Network access only while acquiring full-resolution public GIS data; validation and the committed sample run offline.
- Source licenses that permit redistribution, with attribution and retrieval metadata.