You shook on 50/50. The filing says 60/40.

You both said 50/50 at the kitchen table. You remember the handshake. You remember meaning it.

Then you did the incorporation paperwork alone, late, and typed 60/40. You've been telling yourself you'll square it with them later. The filing went through yesterday. They haven't seen it.

The dilemma

This is a composite — the kind of founder dilemma we see often enough to recognize the shape, assembled from several so no real person is named.

Two founders. Call them you and Maran. A year of circling the same idea, then a weekend where it became real: you'd build the product, Maran would run sales and raise. At the kitchen table, over cold tea, you both said the obvious thing out loud — 50/50, equal partners. You shook on it. Nobody wrote it down. Why would you. You were going to do this together.

Three weeks later you handle the incorporation, because you're the one who's organized about these things. Somewhere between the registered-agent form and the share allocation, a sentence starts running in your head: I've done every night this month. Maran hasn't shipped a line. I carry more risk. So you type 60/40. You tell yourself it's temporary. You tell yourself you'll bring it up when things calm down. You tell yourself it's basically fair, and anyway it's easy to amend.

The amendment window is about a week. The conversation you'd need to have isn't one you've scheduled. And the real question — the one under the spreadsheet — is whether you'd read the filing aloud to Maran's face, today, in those words. You keep not opening the doc. That not-opening is doing a lot of the deciding for you.

The read

We walk this through a few of the eight lenses that actually bite here. Not all eight — the ones that draw blood on this specific gap between the handshake and the filing.

Duty — svadharma: what did you actually commit to?

The duty lens asks a narrow question: what is the thing you took on, in its own plain terms? Not what you now wish you'd negotiated. What you did.

Here, the commitment is unusually clean. You said 50/50 and shook on it. There was no "we'll see how the work shakes out," no "let's revisit in six months." The handshake was the term. Svadharma isn't the role you'd design with hindsight; it's the one you stepped into with your word. By that reading, the 60/40 filing isn't a renegotiation. It's a term you changed while the other party was looking the other way. The duty lens pulls hard toward: the agreement is 50/50, and the paperwork currently lies about it.

Discernment — viveka: is "I worked harder" a reason or a story?

Viveka is the lens that separates what's true from what's merely useful to believe. And it has a sharp thing to say about I've done more nights.

Maybe you have. That can be completely true and still be the wrong frame. You and Maran didn't split equity on a nights-logged basis — you split it on a bet that both of you are load-bearing over years. Your front-loaded nights are real. So is the fact that Maran's work (the raise, the first customers) hasn't arrived yet because it comes later in the build. The discernment lens notices that "I worked harder" arrived after you'd already started wanting more equity, not before. That's the tell. A reason you find before you want the outcome is a reason. A reason that shows up precisely to justify the outcome you were already reaching for is a story. Viveka pulls toward: audit the timing of your own argument.

Non-harm: weigh the cost if you're wrong and this surfaces in two years.

Non-harm doesn't just ask what hurts now. It asks what the action plants.

Right now, nobody's bleeding. Maran doesn't know. The company runs. But run the tape forward. In two years you've raised, you're both exhausted and bonded by the trenches — and Maran pulls the cap table for a due-diligence request and sees 60/40 with a filing date three weeks after the kitchen table. The injury isn't the ten points. It's the timestamp. It tells Maran exactly when you decided their trust was something you could edit in private. That's the harm the lens sees, and it's not reversible the way the share allocation is.

Here's where two lenses openly disagree — and it's worth naming. Discernment might grant you that 60/40 could be the correct long-run split for how this company will actually get built. Front-loaded risk is real; some founders do earn a premium. Fine. But non-harm and duty don't care whether 60/40 is defensible. They care that you installed it unilaterally, after shaking on something else, in a document you're avoiding showing the person it affects. The split might be arguable. The method is not. And a fair number reached by an unfair route arrives poisoned.

The verdict: File the correction to 50/50 this week, within the amendment window, before any conversation — then tell Maran what you did and that you've fixed it; if you genuinely believe the split should be different, you earn that in an open negotiation where Maran can say no, not in a form Maran never sees.

Notice what that verdict refuses to do. It won't let you keep 60/40 while planning to "bring it up." The planning is the trap. The duty lens on founder calls like this is unsentimental: you square the paperwork with the handshake first, and then you're allowed to argue for more — out loud, at risk of hearing no.

Why this isn't a chatbot telling you what you want to hear

Lean toward keeping the split and a general AI chatbot will build you a tidy case for founder risk premiums. Lean toward confessing and the same chatbot will praise your integrity. That's not it reading your situation — it's the documented habit of agreeing with whoever's typing. KarmaLens works the other way: it asks the facts that would move the call (was there a revisit clause? did Maran decline paid work to be here? is the window still open?), makes one committed call, and then doesn't renegotiate that call because you flinched at it. It lays the honest options beside the verdict with what each one costs you — including the option where you keep 60/40 and name its real price. And when a verse fits, the system quotes it verbatim rather than letting anyone paraphrase scripture into a shape that flatters the decision. The point of a committed call is that it can hold still long enough for you to argue with it.

The read-it-aloud test (use it tonight, free)

You don't need us for this one. Take the exact words of the thing you did quietly — the filing, the number, the move you're avoiding showing someone.

1. Write the fact in one flat sentence, in first person: "I filed 60/40 three weeks after we shook on 50/50, and I haven't told you."
2. Now imagine reading that sentence aloud, to their face, today. Not the softened version. That one.
3. Finish this: "I'd rather bring this up 'later' because saying it now would mean ___."

Read the blank. If it names a fact that genuinely isn't settled yet — "we never actually finalized, I jumped the gun" — you have a renegotiation to reopen, cleanly, in the open. If it names their reaction — "they'd be hurt," "they'd argue," "I'd have to give the points back" — then "later" isn't timing. It's the quiet you need to keep the move alive. The flinch at reading it aloud is the verdict. A term you can't say to the partner's face isn't a term you've agreed; it's a term you've taken.

The amendment window closes in about a week — that clock is real and it is short. But the thing it protects isn't the share allocation; the deeper window closed the moment you typed the number and didn't open the doc again. If you want to pressure-test the actual facts — the revisit clause you can't quite remember, whether Maran walked away from paid work to be here, whether the filing is still inside the window — start a founder reading and watch a committed call get made before the story hardens. You can read others like it in the gallery. Here is the last question to carry: if the split you filed is as fair as you say, what exactly are you protecting by not letting Maran see it?