Attackers already use free AI to draft phishing at scale, clone voices from 3 seconds of audio, and triage who’s softest. Every case on this board hands you a real, free AI capability of your own — not a lecture — and never touches anyone else’s accounts.
Change of season changes the work: heating and winter travel, renewals and enrolment windows, insurance and certification paperwork. The showcases below lead with the cases that move with the calendar.
Season is read from your calendar date, not from metadata — the cases never change with it.
Exact catalog counts. Bar length is relative to the largest category, not a risk score.
Every figure here is counted from metadata.json when this page loads: the paid alternative each case prices, what it cites as the cost, and the documented $0 path beside it. Nothing on this panel is typed in by hand. This sum is not a bill you owe.
Sources and dates: each case carries its own cited price and retrieval date in metadata.json under cost_model; the research file behind the price bands is doc/research/affordability-research-2026-10-03.md.
Newest case by publish order
These failed a live spot-check and are held from publication until tool terms or evidence are re-confirmed. Kept visible here for integrity, not hidden.
Reading the catalog…
Every card is a companion app already shipped in this repository — the card links are generated from metadata.json, so a card can never point at an app the catalog does not list. Preview images are decorative mockups, not screenshots or product claims.
Each case should cite a dated source beside the claim. Withdrawn citations say so. Check the source.
Copy-paste prompt for a named free tool — that's the leverage.
Every centerpiece analyzes only your accounts, devices, systems.
Free check is complete alone; paid app only where ongoing monitoring pays.