A review's findings sit with the argument, not inside its conclusion

The two chips a review produces — what the conclusion came out as, and what the argument itself did — used to be drawn beside the conclusion premise, which read as if they were findings about that one premise. They are findings about the whole argument, so they now sit on the argument's own line, next to its premise and claim counts. A draft shows no chips: there is nothing to report until it is published.

Marks where marks were missing

Claim Details, Citation Details and Premise Details now carry the same icons the sidebar's own button strip already used for them — they were the only three sections showing a bare title. The chip on a conclusion premise and the propositional-variable chip in Claim Details each pick up their mark too.

The "Toolbox" heading above the sidebar is gone. It named nothing you needed and took a line of space from the sections beneath it.

A review's two findings now read at a glance

The argument view shows two chips: what the conclusion came out as, and what the argument itself did. The second used to be a sentence — "Doesn't reach its conclusion" — beside the first's single word. It is now a word or two, like its neighbour: Reaches, Falls short, Contradictory.

Each chip also carries its own mark now, so the two are told apart by shape and not only by wording: a checked document for the conclusion, two texts side by side for the argument. The mobile app shows the same pair.

The shorter wording keeps every promise the longer wording made. A review still says what happened rather than grading the author — "reaches" can stand beside a premise you rejected, and falling short is usually just something you left unsettled, which you can go back and settle.

Warning turns orange

The amber used for warnings is now an orange, in light and dark mode alike. It reads as caution on a canvas that already spends yellow-green on the main action, and on the review chips it puts "falls short" clearly between the green of an argument that reaches its conclusion and the red of one whose premises contradict each other.

Agree and disagree, in the same words everywhere

Reacting to a claim now offers Agree and Disagree where it used to say Support and Counter — the words the mobile app already uses, and the ones that say what you are reporting: your own position, not a verdict on the claim.

The three buttons also sit in a new order: agree, unsure, disagree. Unsure is no longer parked after the two it sits between, so the row now reads from one position through the middle to the other.

Two small wordings around a review are tidied up at the same time. The note on a value you did not set yourself now reads as one sentence, and says a claim the argument still has to reach starts out unknown. And sharing a review on an unpublished argument explains the same thing whichever way you meet it: publish the argument first, so there is something for a shared review to point at.

Each grade's explanation is written in that grade's colour

Under the two results at the end of a review, the explanation of each one now names it in the same colour the grade itself is written in — so "Falls short" below reads in the same orange as Falls short above it, and it is obvious at a glance which explanation belongs to which result rather than having to match the words up.

The results details no longer list supporting premises. They were shown without anything you could do to them, so the list reported a state rather than offering a decision. Your assignments, the conclusion and any constraints are all still there.