Most idea validators have the same problem: they want to tell you yes. Paste in your concept, get a 78/100 score with a rocket emoji, and off you go to build for three months on pure optimism. WhittleOS takes the opposite stance. It is deliberately kill-biased, built by solo founder Boris Binyaminov to find the reason an idea should die before you spend a single week on it.
What makes it genuinely different is that it does not wait for you to bring an idea. Point it at a market, or nothing at all, and it goes searching for candidates itself. It reads real public pages, pulls out documented problems with source links, and then throws most of what it finds straight in the bin. That is not a bug. That is the whole point.
On a single published sweep run on 2026-06-04, WhittleOS searched 30 markets, turned up 3,430 candidates, and shortlisted 93. You can see exactly how that filtering works, with no signup required, at the sample Discovery run.
What WhittleOS Actually Does: Key Features
Autonomous Candidate Sourcing
You do not have to arrive with an idea. WhittleOS splits your market into sub-markets, runs live searches across real public sources, reads the pages it is allowed to read, and pulls out candidate ideas with documented problems behind them. Each problem either links to the page it was read on or is labelled as the tool’s own estimate. No invented citations. On the sample run, 60 searches and 24 full pages were read before the first idea was written up.
A Kill-Biased Gate That Returns Honest Zero
Most validators skim toward encouragement. WhittleOS runs every candidate through deal-breaker checks first. Thin market, no documented pain, too much compliance overhead for one person: any of those kills the idea and names the reason. When a market is genuinely weak, the tool hands back nothing rather than manufacturing a fifth idea to fill the page. That is a deliberate design choice, not a failure state.
Code-Computed Scoring Against Your Founder Profile
The 0-100 score is not the language model’s own number. A fixed weighted rubric is set in code; the model rates each area and never sets the total. That means the score cannot talk itself into a pass. It is also tied to your specific situation: your time, budget, skills, and sales style all feed into the verdict. A funded team and a solo founder can get completely different results on the same idea.
Single Idea Check Mode
Already have an idea? You do not need to run a full Discovery sweep. The Single Idea Check puts one concept through the same deal-breaker filters and weighted rubric, returns a Build it, Test it first, or Drop it verdict, and shows you the biggest unknown and one concrete next move. The sample report on InvoiceRescue, published unedited on the site, scored 73 and came back as “Try a paid pilot” rather than a green light.
Marcus at 10pm, One Week Before He Was Going to Start Building
Marcus is a freelance developer in his mid-thirties. He had been sitting on an idea for a solo real estate agent CRM for about two months. At 10pm on a Tuesday, instead of opening a code editor, he ran a WhittleOS Discovery sweep on real estate agent and property management tools.
The tool searched 20 sub-markets, considered 69 candidates, and killed 36 of them, each with a named reason. The top result, a deal pipeline tracker called SoloPipe, scored 52 and came back as “Test it first,” not “Build it.” The verdict cited weak distribution, no verified revenue proof, and a specific walk-away condition: fewer than 30 waitlist signups after 45 days across organic channels.
Marcus told me that was the moment the tool earned his respect. It did not say his market was hopeless. It said: here is what you need to see before you write a line of code, and here is exactly when to quit. He spent that week building a narrow landing page instead of an app. No three-month detour. One week of cheap testing, with a clear exit condition already written down.
How a WhittleOS Discovery Run Actually Works
Step 1: Market Planning
WhittleOS splits your market into the sub-markets worth searching separately, so a broad input like “tools for solo service businesses” becomes a structured set of targeted searches rather than one vague query.
Step 2: Live Search and Page Reading
It runs real searches, keeps the results, and reads full pages where it is permitted. It never follows a link the model invented rather than found. Evidence is labelled by type: read from a real page, seen in search results, provided by you, or flagged as unverified.
Step 3: Deal-Breaker Filtering
Candidates without a documented problem behind them are dropped immediately. Then a second pass removes anything that fails harder checks: too much compliance overhead, no repeatable way to reach buyers, just an AI wrapper with no real product underneath.
Step 4: Scoring and Shortlisting
What survives gets rated area by area against your founder profile. Code turns those ratings into the final score. At most five ideas come out, each with a verdict, linked evidence, and one specific first move to test demand this week. A run takes 10 to 40 minutes per market; you get an email when it finishes.
Pricing: Pay Per Decision, Not Per Month
WhittleOS uses prepaid credits that never expire. There are no subscriptions and no sales calls. You can see the exact credit cost of any run before you spend it, and a failed run refunds the credit automatically. Unused packs are refundable within 14 days.
- Starter: $19 for 2 credits, one market end-to-end or 2 single-idea checks
- Founder Discovery Pass: $49 for 6 credits, five markets or one market plus follow-up deep-dives
- Multi-Market: $99 for 15 credits
- Working Founder: $149 for 30 credits, covers all 29 markets in the catalogue
New accounts get 2 free idea-check credits on signup. No card required to start.
Who Should Try WhittleOS
If you are a solo or indie founder who wants a blunt, evidence-backed answer before committing months to something, this is worth a serious look. It is not for teams chasing funded-startup scale. It is for one person with limited time who needs to know whether an idea is worth testing, worth building, or worth dropping right now. Start with the free sample run at whittleos.com/sample-discovery, no account needed, and see what a real kill-biased output actually looks like.
