the bench

One bench. Every founder decision.

Not prompts, not pep talks. Each tool runs your situation against a defined judgment standard and hands back the read a veteran cofounder would give you: what to fix, what to skip, when to stop.

Prefer answers to a specific question instead?

which founder are you?
the whole bench

Idea to exit. Or idea to income.

your goal. the bench serves both
  1. Is it real?building what nobody wantedroughly here
  2. What do I build?scoping past the pointroughly here
  3. Who finds it?silence after launchroughly here
  4. Will they pay?undercharging on a guessroughly here
  5. Keep going?not knowing what is actually stuckroughly here
  6. exitraise, scale, sellroughly hereincomerun it, keep it, own itroughly here

Every founder meets these five questions. Only the ending differs, which is why the line above names two. Nothing below is locked to a step, so start anywhere: what carries between tools is your own context, and what one read learns the next already knows.

Copy Sharpener
3 rewrites, punchier and clearer.
Customer Discovery Accelerator
Who to talk to, and where they are.
DD Prep
Every likely question, answered first.
First-Channel Finder
One channel, two weeks, a kill number.
Idea Finder
Directions from your own ground.
Idea Validator
4 dimensions, one kill assumption.
Landing Teardown
Every fail gets a rewrite.
Launch Checklist
Tailored to you, gaps first.
MVP Scoper
The one loop that proves it.
One-Pager
One page, two audiences.
Pivot / Persevere Diagnostic
The verdict, read from your numbers.
Pricing Fixer
One number, comps cited, one test.
Problem Sharpener
One sentence, checkable before you build.
Roast My Idea
The fatal flaw, named.
Stack Picker
One opinionated stack, not a table.
Term Sheet Explainer
Plain language. Never tells you to sign.
User Interview Script Generator
The exact questions, in order.

Radars and early access

early access
Competitor Radar
The one competitor move that changes your plan
early access
Fundraising Radar
One investor who fits your round, or nothing
Idea Radar
A startup idea that fits you, found while you work
early access
Pitch Stop
Find the slide that loses the investor, before you send
start from the decision

What is expensive to get wrong right now?

not sure? start anywhere
I'm not sure the idea deserves more time.Roast My IdeaIdea ValidatorProblem Sharpener
I need evidence from real users, not compliments.Customer Discovery AcceleratorUser Interview Script Generator
I'm building too much, or too vaguely.MVP ScoperStack Picker
I'm walking into a fundraising room.One-PagerDD PrepTerm Sheet Explainer
The numbers moved and I can't read them.Pivot / Persevere DiagnosticPricing Fixer
the fundraising desk

Your highest-stakes week, one connected bench.

Runs share context. The one-pager's read carries into DD prep, and DD prep's into the term sheet.

sample reads from real runs

This is what a
read sounds like

nothing invented, nothing softened
Roast My IdeaSharpen it
Illustration: a founder writing a startup idea in a notebook at night, a mug of coffee beside themillustration

Verdict first.

swipe for the read →
the situation

An indie dev building for indie devs.

Can ship it solo. Testing before building.

what nommi said

You're an indie dev building for indie devs, you can ship this solo, and you're trying to validate before you build, which most people skip…

the verdict

Sharpen it.

see the full run →
Landing Teardown
Illustration: a founder at his desk late at night, arms folded, frowning at a landing page that gets visitors but no signupsillustration

Find the leak.

swipe for the read →
the situation

A landing page already live.

620 visitors. 5 signups.

what nommi said

620 visitors and 5 signups feels like the market telling you the idea is dead; it is not…

the read

Not dead. Leaking.

see the full run →
Term Sheet Explainer
Illustration: a founder at a kitchen table hesitating over his first term sheet, the pen still untouchedillustration

Read it before you sign.

swipe for the read →
the situation

A founder's first term sheet.

Read once. Looked normal.

what nommi said

This is your first term sheet and you have read it as normal, which is exactly the trap a first sheet sets…

before you sign

"Normal" is the trap.

see the full run →

19%

slower, measured. Experienced developers using AI believed it made them 20% faster.

METR randomised trial, 16 experienced developers, 246 tasks, July 2025
Start free →

start free · no card