Proof Rack - test a launch page's claims before you back it

proof-rack.skillsafe.ai

Clean

Before you back a crowdfunding campaign or preorder a new gadget, paste the launch page. Free in the browser: the claim finder (every sentence with a figure, a comparison, a test, a demo, a render, hype or health words, marked with what it shows), the page's own quick reading of each claim, a rack of test tubes - liquid height is the rung of evidence, colour is the flaw - and every export (questions for the creator, Markdown, claims CSV, JSON). One metered lane audits the claims: each gets a rung (only said, shown working, measured by the makers, checked by outsiders), the one flaw that weakens it most (nothing real behind it, an unnamed comparison, the best case shown as typical, a test too small) and a question to send the creator, with a verdict of walk away, ask first, holds up or needs real claims. The page re-checks every reply against the verdict rules, renders called proof, numbers the claims never state and accusing words. It judges evidence, never intent, and is not advice to buy. Built on @wanshuiyin/experiment-audit.

proof-rack.skillsafe.ai
PricingUsage-based + 10% creator margin
Billed model rate$2.20 in / $13.20 out per 1M tokens
Creator margin+10%
Effective rate$2.40 in / $14.40 out per 1M tokens
Security scanClean — skill and frontend scanned
Model gpt-terra
Created2026-10-21
Updated2026-10-23

Every public app is built from a security-scanned skill and must pass a clean scan — skill and frontend — before it can be listed. Have a skill of your own? Turn it into an app — or read the step-by-step walkthrough.

Share