How The Index tests, and who publishes it

sites.llc is published by LABS, the studio that builds labs.llc. Each entry in this index is one of LABS’s own desks — how the index handles that conflict.

What that means

We are not independent, and nothing here should be read as if we were. The people who wrote these entries work for the studio that built the tools. A catalogue like this is only worth reading if its rules make that bias visible and hard to act on, so these are ours:

Catalogued for completeness: sites.llc runs Google Analytics 4 under a measurement ID of its own, which places _ga cookies to count visits.

• Evidence for every claim about a labs.llc tool. Each one points to the tool's source in build 537, by file and line, or to a measurement with the minute it was taken.
• At least one real shortcoming per entry. Every entry lists four. Where a comparator is better, the entry says so first.
• Comparators read, not remembered. Every statement about another tool was checked on that tool's own pages on 27 September 2026, between 20:05 and 20:08 UTC; four points were re-checked at 20:46–20:47 UTC during a second review, and those entries give the time. Where a direct fetch failed (Canva refused it, BandLab and Google Earth returned almost nothing, Citi Bike timed out), and for timeanddate.com, we relied on a search of that company's own pages, and the entry says so.
• No stars, no scores, no ratings markup. We publish no star ratings and give search engines no review or rating data. The verdict is a filing note: a sentence saying what a tool is for.
• Nothing invented. No testimonials, no user counts, no quotes that were not said, and no prices we did not see.

How the ten were chosen

Ten tools were drawn at random from a pool of 100 labs.llc tools at 19:55 UTC on 27 September 2026. The seed is recorded so the draw can be repeated: fd6b2f963dcb00c8fc93571a74a17c79. We did not pick favourites, which is why the catalogue runs from a bike-share board to an Easter calculator.

How we test

We test against a local copy of labs.llc build 537, which is byte-for-byte the build that was live, so that testing did not add load to the public site. For each tool we read its source, then made the same requests the page makes: its own page, its relay where it has one, and the public data sources it names. Each request is logged with its UTC time, status, size and duration. Where a tool computes something (a full moon, an Easter date) we ran its own functions outside the browser and compared the answers with an authority.

Some things could not be tested this way, and the entries say so. The music libraries and the Product Lab plates are not in the local copy, so those claims rest on the code.

How to read an entry

Entries are filed A to Z and numbered IX-01 to IX-10 in that order. Each has the same seven parts: what it is, how it works, the test, the comparison, the shortcomings, the filing note and the spec sheet. Screenshots are of labs.llc pages only, captured from the local copy; we show no other company's screens or logos.

The tags on each card mean this:

Live feeds

The tool reads a data source while you use it.

Computed on the page

Some of its answers are calculated in the page itself, with no data source.

No relay

Your browser does the work and talks to the sources itself; no relay sits in between.

Through a relay

Requests pass through a relay, either the site's own or a public one, because the source will not answer a browser.

Exports files

You can save a file from it (CSV or PNG).

Maps & earth

Maps, the sea, bikes and places.

Time & dates

Countdowns, calendars and dates.

Markets & news

Prices and headlines.

Sound & image

Music and images.

Web & domains

Domains, certificates and the plumbing of the web.

Corrections

If a fact here is wrong, or a comparator has changed since we read it, write to labs@labs.llc. We will correct the entry and note the date.