Questions

Practical

Is it really free?

The Terminal is free and open, with no account and no cookies. That is not a trial. The observation layer stays free because it is also the evidence for everything else — a paywalled accuracy record would be worth nothing.

Do I need an account?

No. Watchlists, layout and alert rules live in your browser. Nothing about you is stored server-side.

Can I use the data in my own work?

Yes, subject to the licence of the underlying source, which is shown on every record. Any filtered view exports as CSV or JSON, and there is a public API.

How current is it?

Most feeds are re-read every few minutes. Some publishers release days or weeks behind, and those records carry their lag visibly.

Reliability

How do I know any of this is accurate?

Every record names its publisher and timestamp, so any single claim can be checked at source. For the aggregate claims, the backtest publishes how often signals preceded confirmed events, scored against two null baselines, with the misses shown.

Why does it sometimes refuse to give a number?

Because the sample was too small to support one. A rate computed from six cases is not a rate, and printing it anyway would be the dishonest option.

Is it biased?

Its coverage inherits the biases of public reporting, which over-covers wealthy and open societies. The evidence tiers make that visible rather than hiding it: you can see how much of a picture rests on press reporting versus instruments.

What it cannot do

Does it predict?

Not yet. The Terminal reports and the signals flag anomalies; anything model-derived is labelled. The prediction layer is in development and will ship with an accuracy record.

Can I rely on it for safety or investment decisions?

No. It is an analytical tool, not advice. For anything where money or a person's safety is at stake, use official sources and your own judgement.

If nothing is shown, did nothing happen?

No — it means nothing was reported by a source we read. Absence is never evidence here.