Data Playground / In the lab
Tool · free

Lock your prediction.

Write it down today and prove later that you called it: the match, the election, the price, the office bet. Your text is sealed in your own browser and never leaves it. You publish the seal; the text stays with you until you reveal it.

  • Sealed in your browser
  • Nothing sent or stored
  • Anyone can check it, with or without us
1 · SEAL · stays in your browserREADY
0 / 2000 · Be specific: what, how much, by when, according to which source.
Only used in the text you publish. It is not part of the seal.
2 · REVEAL & VERIFY · for you or anyoneREADY
· it is read in your browser, not uploaded.
Copy it from the original post or message, not from the receipt.
01

How the seal works

The same idea we use for every forecast we test: lock it before the outcome, so nobody can change it afterwards.

THE SEAL

A fingerprint of your text

SHA-256 turns your text into 64 characters. Change one letter, one comma or one line break, and the seal changes completely. Nobody can work backwards from the seal to the text.

THE KEY

Stops anyone guessing

“Denmark wins” is easy to guess and check. So a random secret key is sealed together with your text. Only the receipt has it, so nobody can test guesses against your seal before you reveal it.

THE TIME

Comes from where you publish

Your device’s clock proves nothing. What proves when you called it is a dated post, email or message with the seal in it. We deliberately store nothing, so the proof is yours and doesn’t depend on us.

CHECK IT WITHOUT US

Two lines in any terminal

The seal is plain SHA-256 of a fixed prefix, the key and the text, separated by line breaks. The text has its line breaks as \n and surrounding spaces removed.

KEY=0123456789abcdef0123456789abcdef
printf 'kvantix-seal-v1\n%s\n%s' "$KEY" "$TEXT" | sha256sum

If the output equals the published seal, the text is exactly what was sealed.

HONEST ABOUT LIMITS

What a seal can’t do

It proves what you wrote, not that it was your only prediction. Someone who seals ten different guesses and reveals the one that came true has proven nothing. That’s why a real track record shows every call, including the misses.

That rule is the whole idea behind our Forward Test for signals and forecasts.