A Manifest V3 Chrome extension that finds supported geospatial dataset links on the current page, reads back the geospatial service requests its interactive maps have already made, and opens selected data in GeoLibre.
The published extension is on the Chrome Web Store:
- Open
chrome://extensions. - Enable Developer mode.
- Choose Load unpacked.
- Select this
extensions/geolibre-chromedirectory.
Everything happens after you click the toolbar icon: the extension scans the
document's links, and reads each frame's Resource Timing buffer to recognize the
services its maps requested. It holds no permission beyond activeTab and
scripting, runs no background service worker, and stores nothing.
From the repository root, run:
npm run build:chrome-extensionThe versioned upload ZIP is written to dist/. Store listing copy and
permission explanations are in STORE_LISTING.md; the publishable privacy
policy source is in PRIVACY.md. The policy must be hosted at a public URL
before submitting the listing.
The scanner recognizes GeoJSON/JSON, GeoParquet/Parquet, PMTiles, GeoTIFF/COG,
and ZIP links in anchors, linked resources, selected data attributes, and
schema.org JSON-LD contentUrl/downloadUrl fields. Existing GeoLibre deep
links are unpacked, and a neighboring name.style.json or
name.geolibre.style.json link is paired with the matching dataset.
Source Cooperative repository pages receive additional handling: the extension
reads the complete embedded file inventory even when the visible table is
virtualized, canonicalizes links to data.source.coop, and removes duplicate
page/download links. The popup can filter discovered files by vector or raster
type without changing the current selection.
A map fetches its tiles and service documents from JavaScript, so they are never
links in the page. What the extension reads instead is performance .getEntriesByType("resource"), the record of its own requests that every
document keeps, collected from the top frame and each frame below it. Recognized
are WMS, WMTS, WFS, OGC API Features, ArcGIS Feature Services, XYZ/TMS image
tiles, and PBF/MVT vector tiles. Tile requests are collapsed into reusable
{z}/{x}/{y} templates, and repeated requests from the same service appear once.
Two consequences of reading the buffer rather than watching the network:
-
Worker requests are invisible. MapLibre and similar renderers fetch vector tiles from a web worker, which records them in the worker's own timeline, not the document's. Such a tileset is recovered from the metadata the main thread did fetch: its TileJSON, or failing that the style document, which GeoLibre can resolve a layer from on its own. A style is only offered in its own right when no tileset from its origin was found.
A TileJSON is recognized by name (
tile.json,tiles.json,tilejson.json), which is narrower than every way a server can name one: tileserver-gl serves/data/<id>.json, and that tileset is reached through its style rather than its metadata. Nothing reads the response, so a TileJSON describing raster tiles is indistinguishable from a vector one and is offered as a vector tileset. Such a false positive cannot become a layer: Add Data resolves the document on submit and refuses it when no source layers come out. Reading the body here would mean fetching a cross-origin URL, which needs the host permissions this design exists to avoid.A style stands as a candidate of its own only when its path names it one:
…/style.json,…/styles.json, or an ArcGIS…/resources/styles/<name>.json. The looser…/styles/<name>.jsonis an ordinary theme or configuration route as well, so a style matched that way is still trusted to explain a tileset found at its origin, but never offered on its own, where a page's theme file would appear as a layer. -
The buffer is finite. It holds 250 entries per document by default and stops recording once full. A map's own early requests are normally well inside that, but a very busy page can lose a service added late. Raising the limit needs a
document_startscript, which needs the broad host permissions this design exists to avoid, so the cap is accepted.
A service endpoint on its own is rarely enough to add a layer, so each result
also carries what the page asked that service for: the WMS LAYERS value, the
WFS typeName, the WMTS layer, and — for a vector tileset, whose source layers
live in its style rather than its URL — the style document the page loaded from
the same origin. These arrive in GeoLibre as serviceLayer and serviceStyle
and land in the matching form fields, so the dialog opens ready to submit rather
than on an endpoint with an empty layer field. Because one endpoint can serve
many layers, results are listed per layer (WMS service: topp:states) instead of
collapsing into a single entry per URL.
After loading or reloading the unpacked extension, open one of these websites in a new tab, let its map finish drawing, then open the extension and confirm it lists the expected service. Selecting the result opens the matching GeoLibre Add Data dialog with the service URL filled in. Each row below is a live third-party map, so what it detects is what a real page actually requests.
| Service | Website | Detected service | Layer carried over |
|---|---|---|---|
| WMS | OpenLayers "Image WMS" example | https://ahocevar.com/geoserver/wms |
topp:states |
| WMTS | OpenLayers "WMTS" example | The GetTile request as a tile template |
sgmc2 |
| WFS | OpenLayers "WFS" example | https://ahocevar.com/geoserver/wfs |
osm:water_areas |
| OGC API Features | pygeoapi lakes collection | https://demo.pygeoapi.io/master/collections/lakes/items |
— |
| ArcGIS Feature Service | OpenLayers "Vector ESRI" example | The …/FeatureServer/0 layer URL |
0 |
| XYZ raster tiles | openstreetmap.org | https://tile.openstreetmap.org/{z}/{x}/{y}.png |
— |
Vector tiles, in an iframe |
MapLibre "Display a map" example | https://demotiles.maplibre.org/tiles/tiles.json |
style …/style.json |
Each row above adds a layer that draws, with no further typing: that is the bar for this table. A row that opens the dialog but leaves a required field empty is a bug, not an expected extra step.
The MapLibre row is worth keeping in the set: the map runs inside an iframe, so
it covers services a page reaches only through an embedded frame, and it renders
through a worker, so it covers the tileset recovered from its TileJSON rather
than from a tile request. It also carries a style whose glyph ranges are served
as .pbf; those are fonts, not a tileset, and must not be offered.
Opening a service URL directly is detected too — the response is the page, so
a WMS GetCapabilities document
lists https://ows.terrestris.de/osm/service.
A document's buffer is its own and is discarded when the tab navigates, so a page
with no services (https://example.com) comes up empty rather than inheriting
the page before it.
After selecting a result, add the layer in GeoLibre and verify that the browser console does not report a CORS error while fetching the service.
Remote servers must allow GeoLibre to fetch the selected URLs through CORS.
Complete HTTP(S) URLs, including signed query parameters, are forwarded to
GeoLibre. Cookies and other browser-session credentials are not forwarded, so
cookie-bound or session-authenticated links may fail. Temporary blob: and
browser-internal links cannot be transferred.