Build an interactive web map with AI
Start with one verified location or boundary, choose the opening view and keep the map files editable in your own project.
You can build an interactive web map with AI around one clear question, such as where a visitor should enter a park. Give map-init the gate's verified coordinates and decide how much surrounding street detail the opening view should show. The skill sets up the page, styling and map script, then adds a first layer you can inspect in a browser.
A point marks a location. A boundary communicates a different claim about the space. If you need the latter, supply a polygon you can use or choose a source with suitable reuse terms. An approximate point for testing can stay explicitly labelled in the code until a verified replacement is available.
What it does
- Show an event entrance at a useful initial zoom level.
- Place an authorised boundary over a quiet basemap so its outline remains legible.
- Check that incoming coordinates put longitude before latitude.
- Keep the map's source attribution visible when changing the visual layout.
- Identify when a large proposed layer needs a tiled delivery format instead of one large file.
Install and use
- Use a project folder, browser and local web server. The default MapLibre and OpenFreeMap setup needs internet for the basemap and no access key.
- Give the agent a centre, zoom and first data layer, with its source and any reuse conditions.
- Open the result locally and check the marker, framing, colour contrast and attribution.
FAQ
Source
Made by Ford Henriod. Original skill: map-init, published on GitHub under the MIT licence.
View on GitHub