Diffrank: a landing page for the AI Code Review Queue idea

Read the 40 lines that matter, not the 900-line diff

People say agents write code faster than they can understand and review it. The headline names that gap with one concrete picture: a huge diff where only a small part needs a human. It promises less reading, not more checking, which is the fix they are asking for.

Words it borrows from real posts:

• “creation cheap and verification expensive”

1. Navigation: It gives tech leads quick jumps to how it works, features, pricing and FAQ, and keeps a trial button in view for anyone ready to act.
2. Hero: It states the promise, shows a sample risk summary sorted into read first, skim and skip, and answers the worry about code privacy before anyone asks.
3. Problem: It plays back what people wrote: review is the bottleneck, and code can pass tests and match style yet still not fit the system. Readers see their own pain first.
4. How it works: Four short steps show the tool only reorders attention and leaves the decision with the reviewer. This matters to people wary of trusting more automated output.
5. Features: Each feature maps to a stated pain: risky lines buried in noise, code that passes tests but doesn't fit, a growing queue, and concern about where code goes.
6. Who it's for: It names the person every agent change waits on and says who it is not for, so the right readers recognise themselves and others leave early.
7. Comparison: It sets ranked triage against the two things people already try, review bots and reading every line, and shows the gap each one leaves.
8. Pricing: It explains that seats are per reviewer, so teams running many agents aren't penalised. That fits people whose agents produce more than they can review.
9. FAQ: It answers the doubts a careful reviewer would raise: cost, code handling, how far to trust a skip label, setup time and platform limits.
10. Final call to action: It restates the core relief, no longer reading every line, and gives one last clear button to start the trial.
11. Footer: It closes with a one-line summary of what the product does, so the page ends on the same idea it opened with.

• Team: $25 a month
• Team Annual: $240 a year

Pricing is per reviewer because review time is the scarce resource, not agent output. A monthly plan lets teams try it with low commitment. The annual plan costs less than twelve monthly payments, for teams that already know they want it.

The button says Start 14-day free trial. After the click: Diffrank is almost ready

Start 14-day free trial asks for a low-risk first step that fits a tool people judge by trying it on real changes. The product doesn't exist yet. The dialog saying it is almost ready lets Ghostlaunch count real interest without promising a working product.

Brand colour #2E6B4E, the “editorial” font pair, a light theme.

A muted green on a light theme reads calm and careful, which suits a tool about risk and trust. The editorial font pair makes the page feel like something to read carefully, matching people who review code closely and are tired of noisy output.