Built an AI that scores your business idea as LEGIT, PIVOT, QUIT, or SH*T for free. Six characters argue and reach a verdict.
This is for the single mom who's exhausted after work and lying in bed at 11pm wondering if her side hustle idea is any good.
The guy on the bus coming home from a job he hates, scrolling his notes full of business ideas, not sure which one is worth trying.
The person who's googled "is dropshipping still profitable" forty times and gotten forty different opinions.
Spent the last two weeks building Check My Hustle for these people. You put in a business idea. One of six AI characters analyses it and tells you in 30 seconds whether it's LEGIT, PIVOT, QUIT, or SH*T. Real market data, real competitors, real numbers. The five other characters chime in with a one-liner.
The six characters:
- Russell — 1970s street hustler, calls you "baby" and "kid"
- Southside Sal — connected guy from New York
- Dex — Gen Z data analyst, lowkey ngl
- root — anonymous hacker speaking in code
- Victoria — corporate exec, no time wasted
- Lord & Lady Blackwood — old money perspective
It's free. Every verdict. No login. No email signup. No 20-page corporate PDF.
If you get a PIVOT (your idea is close but needs an angle), one dollar — less than a cup of coffee — unlocks the Full Play: who exactly to target (down to the specific subreddit), one real competitor to study, a 7-day action plan, and the one risk that will kill your business. Then it walks you to "Now Let's Build This" with the actual next steps — domain names, hosting, the practical stuff.
Other validators charge $100-200 for the same information dressed up in jargon. I wanted a tool that respects your time and your wallet.
Built solo as a non-coder, with Claude doing the heavy lifting on the code. Cloudflare Workers for hosting, Gumroad for the $1 checkout.
Honest feedback welcome — especially if you put in a weird idea and the verdict doesn't make sense.
The character format is a genuinely sharp wedge, and the $1-vs-$200 contrast lands well.
One structural thing worth naming, since it's what your "feedback when the verdict doesn't make sense" invite is really fishing for: a one-line idea can't contain enough to legitimately separate "QUIT" from "LEGIT." The verdict feels authoritative; the input can't support that. It's baked into the format, not a bug.
But your pricing already solves it. The free verdict is the entertaining hook — six characters arguing is a bit. The $1 Full Play is the real product. So accuracy of the free verdict matters less than it feels like it should. Lean into the bit openly, sell the Full Play. Don't over-defend the loss-leader's rigor.
Curious about the scoring methodology — are the 6 characters trained on specific founder archetypes (investor,
ex-founder, marketer, etc.) or general LLM personas? Asked because I built something AI-adjacent and the personas in
my prompt drift over runs unless I lock them with detailed character sheets.