← The AI Hype Audit — all 100 verdicts

PARTLY TRUE Drop "Prompt" and get the prompts behind "these websites" Facebook · Jul 26, 2026

The claimA reel showcases websites supposedly built with AI prompts. Comment "Prompt" and the page DMs you the link to those prompts. The caption's pitch: AI has made an attempt so cheap — one evening — that "I can't do that" is no longer a valid excuse.

Split verdict on a split post. The philosophy in the caption is the truest thing on this page: prompt-to-website tools genuinely do collapse the cost of trying — a competent person with Claude, v0, or Lovable really can stand up an impressive site in an evening. What can't be verified is the load-bearing implication: that the specific prompts behind the gate produce the specific websites in the video. Reels in this format routinely show cherry-picked takes or, worse, showcase sites that were never prompt-generated at all. And the caption's cut-off last line — "that's the whole idea behind our project" — tells you what the reel actually is: an ad with a philosophy essay stapled to the front.

What holds up

  • Prompt-to-website is a real, mature capability: Claude, v0, Lovable, Bolt all deliver
  • "The price of an attempt is one evening" is genuinely true — and worth internalizing
  • The caption is honest that this promotes a project, even if it never names it

What doesn't

  • No evidence the gated prompts produced the sites shown — this format is cherry-pick central
  • Anonymous page, unnamed "project" — you're DM-ing your engagement to a funnel with no face
  • Comment gates exist to train the algorithm and harvest DMs, not to share knowledge
  • One "master prompt" doesn't replicate a polished site — real results take iteration passes
  • Whatever the link sells arrives only after you're inside the DM automation

The catch

The philosophy is the bait and it's genuinely good bait — cheap attempts are real. But notice what the gate is for: a comment costs the page nothing and earns them reach, your DM thread, and a warm lead for an unnamed "project." If the prompts were the point, they'd be a link in the caption. Verdict would move to High confidence with a screenshot of the websites shown — send one in and we'll re-open it.

How to actually do it

  • Skip the gate. Open Claude (claude.ai, Pro $20/mo) and describe the site like a creative brief, not a wish: business name, who it's for, the one action a visitor should take, 3 sections you want, and 2 example sites whose style you like
  • Ask for a single-file HTML landing page first — one file is easy to preview, host, and iterate on
  • Iterate in plain English 3–5 times: "darker background", "make the headline bigger", "add a testimonial section" — iteration, not the first prompt, is where the polish comes from
  • Free hosting: drag the HTML file into Netlify Drop (app.netlify.com/drop) — live URL in under a minute, $0
  • If you want the heavier route: v0.dev or lovable.dev generate full React sites from the same kind of brief — free tiers exist, paid from ~$20/mo
  • Total honest budget: one evening for a real draft, a second evening to make it good

Graded the checkable claims, flagged the unverifiable one, and handed over the ungated version: how to get a real website out of one evening with AI, no comment required.

Effort
One evening to a live draft site; a second session of iteration to make it client-worthy
Cost
$0–20/mo — Claude Pro or a builder's free tier, plus free Netlify hosting
Stack
Claude (or v0/Lovable/Bolt) + Netlify Drop for hosting — no gated prompt pack required
Confidence
Medium
Posted by
a faceless AI-motivation page running a comment-to-DM funnel for an unnamed "project"

See the original claim →

We test hype for free. We build the real thing for a living.

Thirty minutes, no pitch — and you'll leave with something useful either way.

Book a call with Todd or start with the free Business Checkup →

The Verdict Weekly

Three verdicts every Friday. Free forever, unsubscribe anytime, no spam — that would be ironic.

© Schreier Group · schreiergroup.com · See a wild AI claim? Drop it here and we'll test it.