Does your app actually work offline?

Most say they do. We open yours in a real browser, wait for the service worker, cut the network, and read what is still there. You get the manifest, the cache contents, and the two or three things Android will do differently than you expect.

We store the address and your email to run the check and send the result. Nothing else.

What we check

  • Installable — manifest found and parsed, display mode, icon sizes, and whether a maskable variant exists (the one Android needs to avoid putting your square icon in a white circle).
  • Offline — service worker registered and activated, and how many files are actually in the cache. An empty cache with a registered worker is the most common false positive.
  • Identity — whether id is pinned, so a future change to start_url does not orphan everyone who already installed.

What we do not do

No score out of 100. A score averages things we did not measure, and this directory only publishes what it observed. You get the observations.

Why it is not instant

The measurement drives a real browser — load, wait, disconnect, re-read. It runs in batches and the report arrives by email, usually within a day. We would rather be slow and right than instant and guessing from your HTML.

This is the same measurement behind every one of the 592 apps in the directory, of which 249 keep working with the network off.

What we learn from measuring 592 web apps

One email a month, only when there is something worth sending. Last one: most of the PWAs we measured ship no maskable icon, so Android puts their square logo in a white circle.

One email a month. Only the apps worth your time.

What they do, what they don't, and whether they actually work offline — we check. No roundup of everything we listed, no affiliate links.

Double opt-in — nothing is sent until you confirm. Unsubscribe any time.