daniele.michelin // blog

DE EN IT

~/blog / scam-radar-current-phishing-warnings

Scam radar: today's phishing emails and fake texts in plain words, updated twice a day

1 November 2026 · Daniele Michelin · ~2 min read · Also in DE · IT

Before Black Friday and Christmas, fake texts, phishing emails and fake shops peak. I built a scam radar: one page with the current scams in plain words, each with a tip and its source, updated automatically twice a day.

#security #phishing #fraud #automation #ai #n8n

Topics: AI, LLMs & RAG Security & patches

🎬 Explained in 20 seconds

Scam Radar: current phishing warnings in plain words

daniele@mikweb: ~/blog — less scam-radar-current-phishing-warnings.md

Every year before Black Friday and Christmas it is the same picture: parcels that supposedly cannot be delivered, subscriptions that end “in 72 hours”, fake shops using well-known names. Anyone who works in support knows the calls that follow. That is why I built a scam radar: a page that explains the current scams in plain words so you recognise them before you click.

What you find there

Every entry has the same shape: what the scam is, how it works and what to do. Each one links to its source and has buttons to forward the warning to parents and friends on WhatsApp or Telegram. The page is available in German, English and Italian, with no cookies and no tracking.

The 5 warning signs

Time pressure or threats: “account blocked”, “final notice”, “delivery failed”.
You are asked to tap something: open a link, install an app, enter data.
The sender does not fit: different domain, odd number, spelling mistakes.
It is about money or credentials: no bank asks for passwords, TANs or codes by e-mail, text or phone.
It sounds too good to be true: a prize, a refund, guaranteed crypto profit.

If in doubt: tap nothing, do not call back the number in the message, open the official app or website yourself. Already clicked or entered data? Change the password at once, call your bank, have the card blocked and report it to the police.

How it works behind the scenes

The radar maintains itself. The flow I built in n8n:

n8n (08:00 and 16:00)
   │  reads public warning pages (Watchlist Internet, Mimikama)
   ▼
AI model (IONOS)
   │  picks real scam warnings, writes title, explanation
   │  and tip in German, English and Italian
   ▼
scam-ingest.php on the website
   │  accepts requests only from the n8n server (IP list, no password)
   ▼
Python script builds the pages (DE/EN/IT), the RSS feed and the sitemap

Two rules mattered to me. First, the model may use only facts from the source text and must not invent numbers, names or amounts. Second, every entry shows its source so you can read the original. The intake on the website has no shared password that could sit in plain text somewhere: it only accepts requests from the addresses of the n8n server.

Limits

The selection runs automatically and can contain mistakes. If you find one, write to me and I withdraw the entry. The radar is not legal advice and not an official warning; it is a help to recognise fraud faster.

And if you are unsure?

For a single suspicious message there is the scam check: paste the text, get an assessment in seconds, no sign-up. The radar shows what is going around; the check helps with the message that just reached you.

← All articles