We pointed our newest tool at our own homepage. It called our hero "a category, not a problem." Fair.

Last week we shipped a tool that takes a red pen to landing pages.

The first page we ran through it was ours.

Mark number one, severity critical: "The hero names a category, not a problem." Five seconds in, a visitor knows which shelf we sit on. Not what breaks if they close the tab. The tool was right. It stung anyway.

That tool is called ROAST, and it lives in a new corner of our site called Labs. This post is the tour.

What Labs is

Pulse Browser Inc. is a small team with one flagship product. But building a financial OS teaches you a rule that travels well beyond budgets and brokerage accounts:

A good tool settles a decision.

Not "provides insights." Not "empowers your journey." It takes one question you were going to answer badly, with vibes and a gut feeling, and answers it with numbers you can check.

Labs is the register of every tool we ship. Each one has to pass the same three tests before it earns a number:

•Real data. No mock numbers, no invented benchmarks, ever.
•Math you can inspect. Scores are deterministic. The AI writes sentences. It never grades.
•A price a rational person says yes to.

Here is the register so far.

01. Pulse

The decision it settles: where your money actually stands, today.

The flagship. Connect your accounts once and get the full picture: budgets that roll themselves forward, portfolio analytics with the math shown, a personal brief every morning, and a health score that tells you the one thing to fix next. Web and iOS. Free to start.

02. SaaS X-Ray

The decision it settles: whether that micro-SaaS is worth what they are asking.

People buy small internet businesses the way people used to buy used cars. The seller says "runs great." You squint at a revenue screenshot. You wire the money.

X-Ray is the inspection. Paste a TrustMRR listing and get a 0 to 100 score, a valuation range with the multiple decomposed, risk flags ranked by how much they should scare you, and the exact questions to ask the seller before you wire anything. Coverage gets its own score, so when the data is thin, the report says so instead of guessing.

$39 per report. No account needed. Under a minute.

03. ROAST

The decision it settles: why your landing page is not converting strangers.

Six dimensions: clarity, credibility, differentiation, friction, urgency, proof of value. Every finding is a mark pinned to your actual page, quoting your own copy. No generic advice. The full roast ends with the hero section we would ship instead of yours.

Your first mark is free, no account. The full roast is $19.

And because we believe in eating the cooking: we ran it on levels.io, shipfa.st, and photoai.com. Those reports are public, on the ROAST page, scores and all. Nobody cleared 80.

04. Untitled

The notebook is never empty. Experiment 04 is sketched, argued about, and not yet shipped. It follows the same three tests as the first three.

If you want to hear about it exactly once, when it ships, there is a notify list. One email. Nothing else.

The rule behind the rule

Why does a personal finance company ship a due diligence tool and a landing page critic?

Because they are the same product underneath. Someone is about to make a decision with real money attached: buy this business, keep this homepage, trust this budget. The lazy version of that decision runs on adrenaline and a screenshot. The good version runs on numbers someone is willing to show their work for.

We only know how to build the second kind.

If a tool can't tell you what decision it settles, it isn't a tool. It's a subscription.

We build the other kind.