Public-source fact-checks of security news
Is it a hack?
Here is what is actually known.
When something big breaks, rumors arrive before facts. We sort every claim about an outage, breach or security story into what is confirmed, what is only reported and what has been debunked, and we link each one to its source.
- Gathered and organized by AI agents. They read public sources on a schedule and file what they find.
- A person makes the final call. A case is closed as a hack, or as not a hack, only after the site's owner has reviewed the evidence.
- Every edit is public. The pages, the data behind them and their full history live on GitHub.
Made for entertainment using AI. Not endorsed by anyone named here.
What that meansDon't take our word for it.
We are not asking you to trust us. We are asking you to check. Everything here is built so that you can.
- 01
Every claim links to its source
Each claim shows where it came from and why that source supports it. No source, no claim.
See a sourced page - 02
Every change is logged
Each page carries a dated update log, newest first. Underneath it, GitHub keeps the full edit history.
View the edit history - 03
The data is public
Every page is built from plain data files you can open, download and compare with the sources.
Open the data - 04
Corrections stay visible
When we get something wrong, we fix it and leave a dated note saying what changed. Nothing is edited quietly.
Report an error
Incidents
-
FAA Nationwide Ground Stop: What We Know
- Confirmed 3
- Reported 4
- Unverified 5
- Rumor 3
- Debunked 2
Who does the work
A small desk of AI agents, each with one job, hands work down the line. One person makes the call that closes a case.
- Step 1 · AI agent
Research
Watches public sources on a schedule and logs what it finds, with links.
- Step 2 · AI agent
Verification
Checks each claim against the sourcing rules and proposes a label.
- Step 3 · AI agent
Writing
Turns verified facts into the page, with a source on every claim.
- Step 4 · AI agent
Editor-in-Chief
Reviews the draft against the rulebook, then publishes it.
Above the desk · A person
The final verdict is a person's call
The agents label individual claims as the evidence comes in. The verdict on a whole incident, Confirmed hack or Confirmed not a hack, is published only after the site's owner has reviewed the evidence and approved it. The agents never contact anyone, never log in anywhere and never use leaked data.
Six labels, in plain words
Every claim carries exactly one. The fuller the dot, the firmer the footing.
- Confirmed
- An official document states it, or at least two independent sources report it. Outlets repeating one source count as one.
- Reported
- Sourced journalism that has not been officially confirmed. One outlet, or several relying on the same source.
- Unverified
- Claimed in public, but not yet checked or checkable, or the accounts conflict.
- Rumor
- Circulating with weak or no provenance. Shown clearly as a rumor, never as fact.
- Debunked
- Shown to be false by an official source or independent evidence. The original claim stays visible, with the evidence.
- Not established
- The evidence does not settle it either way. Used for headline questions such as whether something was a cyberattack.
Open source, open history
It all lives on GitHub.
The pages, the data behind them and every edit ever made sit in one public repository. Read it, compare versions, or tell us what we got wrong.
- primeobsession/isitahack-dot-com
- data/incidents/<slug>/
- One folder of data files per incident: claims, sources, timeline, update log
- incidents/<slug>.html and .md
- The page and a plain-text version, both built from that data
- format/
- The definition of every data field, and what each is for
- scripts/validate.mjs
- The checks every incident must pass before it is published
- scripts/build.mjs
- The script that builds these pages, with no dependencies
- index.json, llms.txt
- The list of incidents and a map of the site, for programs and AI assistants