60-second read

Why browser-only processing is different

Most "free" privacy and PDF tools online send your file to a server, do the work there, and ask you to trust them. Today's Tasks doesn't have a server in the loop. Here's what that means β€” and how to verify it yourself.

1. How most cloud PDF / privacy tools work

πŸ“„
Your file
β†—
☁️
Their server
file copied here
↓
βš™οΈ
Processing
on their machines
↩
πŸ“„
Result back
trust they delete it

Even reputable tools advertise that they "automatically delete files after 2 hours". You're being asked to trust that promise. For most documents that's fine. For a contract, a medical record, a tax return, or a confidential settlement β€” it shouldn't be.

2. How Today's Tasks tools work

πŸ“„
Your file
stays on your device
β†’
πŸ–₯️
Your browser tab
processes it locally
β†’
πŸ“„
Result
stays on your device
No server in the loop. 0 bytes uploaded.

Every privacy tool on this site is built with browser-native JavaScript (and where needed, WebAssembly). The file you pick is read by your browser, processed by your browser, and the result is written back to your browser. No network request ever carries your file.

3. Who needs browser-only processing?

Legal

Lawyers & paralegals

Attorney-client privilege is broken the moment a draft contract is uploaded to a third-party server.

Health

Doctors & clinics

HIPAA, GDPR, and patient confidentiality apply even when a tool promises 2-hour deletion.

Finance

Accountants & auditors

Client tax returns, bank statements, and ledgers don't belong on a free tool's servers, ever.

Press

Journalists & activists

Source documents, leaked materials, and case files require zero-server handling β€” not "we'll delete it later".

4. Verify it yourself in 60 seconds

  1. Open a Today's Tasks tool β€” try Document Redactor or Password Generator.
  2. Open DevTools. Press F12 (or ⌘βŒ₯I on Mac), then click the Network tab. Reload the page to start clean.
  3. Use the tool. Type, paste, generate, redact β€” whatever the tool does. Now look at the Network panel.
  4. Count outbound requests carrying your input. You'll find: zero. The only network requests are for the page itself, the CSS, and (where applicable) an anonymous k-hash lookup for breach checks.

You can do this test on any tool on this site, any time. We don't get to fake the Network panel.

5. What you should actually look for in the Network panel

“Count the requests” is the short version. The longer version is worth knowing, because it is the difference between checking and assuming.

With the Network tab open and the page reloaded, use the tool β€” paste text into the Document Redactor, generate a password, encrypt a secret. Then read the list. Requests for the page itself, the stylesheet, the fonts and the ad script will be there; those are the page loading, and they carry nothing of yours. What you are looking for is a request that appears at the moment you acted, and whose method is POST or PUT rather than GET. Click it and read the request body. If your file or your text had been uploaded, it would be sitting there in plain sight, because that is the only way it could reach a server.

There is no version of this where a tool uploads your file and the Network panel stays quiet. The panel is part of your browser, not part of this site β€” which is exactly why it is worth more than any promise on a marketing page.

6. The two places bytes really do leave β€” stated plainly

A page that claims “nothing ever leaves your device” without exception is easier to write than one that is accurate. Two tools on this site do make a network call, and both say so on their own pages:

Everything else β€” redaction, metadata stripping, encryption, the vault, the calculators β€” makes no outbound call carrying your input at all.

Separately, and worth being straight about: the site itself carries third-party advertising and anonymous analytics, described in the privacy policy. Those are about the pages you visit, not the contents of the files you process β€” but “no server sees your document” and “no third party is present on the site” are different claims, and only the first one is ours to make.

7. What the tools do instead of a server

Removing the server does not remove the work; it moves it into the tab. A few concrete examples, each documented on its own tool page:

The trade-off is consistent: you give up cross-device sync and server-side recovery, and you get a tool where “who else can read this?” has a checkable answer rather than a promised one.

Try a tool now β€” no upload, no account