A LITTLE SPACE BEFORE YOU ACT
A moment of pause.
A clearer next step.
Look for clues in a suspicious message.
Understand the pattern. Verify independently.
Clues, not certainty.
This page explains possible patterns.It cannot prove a message is safe.
What did you receive?
Processed in this page. No message storage or upload.
Or explore a made-up example SYNTHETIC
A small starting point. Five patterns, English keyword rules, no live checks. All examples are invented; detection accuracy is unmeasured.
Make sense of it
LOCAL RULESRECOGNITION IS A SKILL
Learn the story behind the request.
Explore five common setups and ways to verify them.
THE PATTERN LIBRARY / 5 EXPLAINERS
Know the setup.
Recognize the pressure.
Defensive education based on official consumer guidance.
A small selected catalog, not a complete or live scam database.
Source review: 10–11 October 2026. FTC sources use a US consumer context; their reporting routes are not worldwide services. These explainers are paraphrases, and the local matching rules are a design proposal.
SHOGO MAKISHIMA TOOLS
An honest prototype.
A broader ambition.
The full intent is a list of scams, how they work, and detection.
This interactive slice explores one possible beginning.
What works here
Paste or type English text; review exact keyword matches; inspect parsed URL hostnames; explore five sourced pattern explainers; build a verification checklist; download a text review without your message.
All processing is local JavaScript. There are no external scripts, analytics, accounts, browser storage or network lookups. Clear removes the current input and result from the page. Refresh resets it to a synthetic sample.
What it cannot establish
The rules cannot authenticate a sender, classify all scams, inspect attachments, listen to calls, read your inbox, inspect a transaction or check a live URL reputation list.
There is no trained model or validated probability score. Precision, recall, false positives and abstention performance have not been measured on an independently labelled dataset. Quotes, negation, ordinary payments, unfamiliar wording and other languages can defeat these rules.
Sources and proposed scope
Audience proposal: a person reviewing a message before responding. English only; educational guidance draws on US FTC sources. No local reporting service is connected.
The supplied baseline and completed research handoff propose pasted-message and URL triage. The handoff arrived during this build and was reviewed before completion. Choosing this small scenario does not replace the full idea.
URL checks are unavailable
Addresses are parsed locally and shown as plain text. No pasted address is opened or sent to a provider. Paths, queries, fragments and user information are omitted from the URL detail display.
Google documents Safe Browsing as a URL threat-list service for non-commercial use, with Web Risk for commercial uses. Neither is connected here. A URL lookup alone would not verify an entire conversation.
DESIGN SIMULATION / NO LIVE INTEGRATION
What would a URL result actually mean?
Explore invented provider states. These fixtures are separate from your message and cannot check a real URL.
“5- Lists of scam and ways to do it
detection”