The Most Useful Site in the World ← Back to the tools
About · Why this exists

About the project

A small site with a large catalogue, built by one person in Luton, UK — and the principles behind the design decisions.

Last updated:

👤 Who built it

The Most Useful Site in the World is the work of Russell Head, a software developer based in Luton, England. He builds and maintains the catalogue independently.

The site is one of the largest deliberately-no-frills collections of browser-based tools on the open web — 1424 calculators, converters, simulators, and visual toys at the time of writing. Every tool runs locally in your browser, with no account, no email, no analytics on the tool page itself.

Russell maintains the project solo. There is no team, no investors, no board, and no company. The domain is paid for out of pocket each year, and hosting is free on GitHub Pages. Donations are optional; every tool remains available without payment.

📊 By the numbers

1424
Tools
29
Categories
12
Languages
0
Logins
0
Ads on tools
CC-BY 4.0
Source licence

💡 Why it exists

The internet used to have a million small useful tools, and most of them were on a personal homepage or a university server. The modern search results for "calculate mortgage" or "what's my BMI" are dominated by sites that exist primarily to harvest an email address, sign you up for a newsletter, or sell your calculator output as a lead to a lender.

There is no business model here. There is no newsletter, no lead form, no upsell, no pop-up, no "download our app", no "create an account to save your results", no "we use cookies for analytics, marketing, personalisation, and partners" panel. The whole point of the site is that the calculation is the product. Once you have your number, the site has done its job and gets out of your way.

That is also why every tool is deliberately available as a standalone URL: https://www.themostusefulsiteintheworld.com/tool/mortgage.html. You can bookmark it, link to it from anywhere, or embed it in another site. The tools do not have to live inside this site to be useful.

📜 The principles

  • Free means free. No paywalls, no "free with sign-up", no "free trial, then £9.99/month". The site is funded (such as it is) by donations, not by gating content.
  • No tracking on the tools. Tool pages do not load analytics, ad-tech pixels, or fingerprinting scripts. Most calculations run locally; a small, documented set of tools makes scoped requests for live data, external APIs, or optional media/audio. This site uses Google Analytics on selected public pages for aggregate traffic measurement; a small, documented set of tools makes scoped requests for live data, external APIs, or optional media/audio, and those exceptions are classified and checked automatically before changes ship.
  • No logins. There is no account system of any kind. "Save your results" is implemented in your browser's local storage, not on a server.
  • Almost everything works offline. Once a tool is loaded, it keeps working without network — that's a feature, not a coincidence, and it's how the site stays useful on poor connections. The exceptions need the network by design (live listings and encyclopedias, spell-check and translation APIs, or optional media/audio playback); each one says so where you use it and fails gracefully offline.
  • Honest numbers. When a calculation has uncertainty, the page says so. Mortgage affordability does not promise "exactly what you can borrow"; a race-pace predictor does not promise a 3% improvement. The form is "here is the calculation; here are the assumptions; here is what would change the answer."
  • Deliberate smallness. One person, one repo, one set of opinions. This means a tool might be exactly what you wanted, or it might not have the feature you expected. That's the trade-off of small-team software — you get honesty, you get speed, you do not get product-management focus groups.
  • Open source by default. Every tool is published under CC-BY 4.0. You may reuse the source for any purpose, with attribution. If you fork it and improve a tool, the upstream site gets nothing, and that's the right outcome — the work is the commons.
  • Product boundaries. The catalogue and its tools stand on their own. Tool pages contain no cross-product advertising or navigation; that boundary is checked automatically before changes ship.

⚙️ How it's built

The published site is static HTML, CSS, and vanilla JavaScript. The public repository also includes scripts to build catalogue indexes and run verification checks; the tools themselves need no framework or account and run in your browser.

The site is hosted on GitHub Pages. The repository is public. Commits, releases, and the full changelog are on github.com/mrpr0phecy/mrpr0phecy.

Architectural decisions are documented in ARCHITECTURE.md (the source of truth for any contributor) and INCOME.md (the honest version of how the project pays for itself).

📅 Recent milestones

🔬 How the tools are tested

Every claim on this page is checked by machine before any change ships — the same checks run on every proposed update, and a failure blocks it. At the time of writing that means:

  • Money maths is tested against the statute, not against itself. An automated suite of 100 checks verifies the tax, National Insurance, student-loan and mortgage tools against independent implementations written from the published rules — including a sweep of every salary from £0 to £200,000, because the bugs that matter live on boundaries. Statutory figures are refreshed every April.
  • Privacy is scanned, not promised. Every tool is scanned for network calls on every change. New tools must make none without an explicit reason; exceptions (live data, spell-check and translation APIs, chart libraries, and optional media/audio) are classified, labelled where you use them, and re-reviewed when touched.
  • Published numbers are derived, never typed. The tool count on this page comes from counting the catalogue at build time — as do the sitemap, the search-engine indexes and the machine-readable catalogue. A number that can't be re-derived doesn't get published.
  • Links and loading are regression-tested. Every internal link must resolve, and the catalogue loader (including retry and scroll behaviour) is covered by executable tests. The full pre-ship check runs the whole gate — including a probe of the live site after every deploy — and the results travel with every change.

The checks live in the open: the repository is public at github.com/mrpr0phecy/mrpr0phecy, so any of this can be verified rather than taken on trust.

📝 Corrections

Spotted a wrong number, a broken tool, or a claim on this site that doesn't survive checking? Report it to hello@themostusefulsiteintheworld.com — one inbox, one human. Money-tool fixes get a new automated test first, so the same bug can't quietly return; everything else is fixed in the open, in the public changelog.

✉️ Contact

For everything: hello@themostusefulsiteintheworld.com. One inbox, one human, typical response in 48 hours.

For press, see the press kit. For security disclosures, see /.well-known/security.txt.

Want to use it?

🔲 Open the catalogue 📚 Full index ⭐ Most popular 🆕 Just added 🎯 By use case ☕ Donate