Added

• Responding takes you to your response. After you add a point to your response, it opens with that point selected and in view. Respond on a claim or a step you have already answered now reads "Go to your response" and takes you to your answer, instead of opening the form again.

• You can go from a response to the words it answers. A response's quotes of the argument it answers have a "Go to" link that opens that argument at the claim or premise being quoted.

• A guide helps you build a representative argument. While you build an argument that represents a source text, and your response to it, both show "Building a Representative Argument": the steps to follow, and a button to switch between the two. Put it away and it stays put away.

• The contents list the Responses section. When an argument shows its Responses section, its contents end with a Responses entry giving how many are published; choosing it takes you to the section.

• Anyone reading an argument can open a claim's or a premise's details from the details button beside it. If you can edit the argument, "View details" is the first entry in the claim's or premise's menu instead.

• A response's own claims are marked R, so you can tell them from the words it quotes from the argument it answers.

• A source text that is a post on X now shows the post as it appears on X. Above the text of an argument built from an X post, you see the post itself — who posted it, their name and picture, when, and anything attached — once you choose View full text on a passage, or open the source text's own reading view. If the post cannot be loaded, a short note says so and the text shows as before. The privacy policy has a new section, "Posts Shown from X", about what X learns when it shows you a post.

• You can report or block an argument's author from the argument's menu. The menu under an argument's title now has Report and Block author when the argument is someone else's.

• Turning on Debug mode in your settings now shows your user ID as the last line of the Account section, ready to copy into a message to support.

• Proposit now tells you when someone responds to your argument, and when an import you started finishes or fails. A bell at the top of every page shows how many are new, and opens the list of them; choosing one takes you to the response or the import.

• You can have these notifications arrive in your browser even when you do not have Proposit open. Turn them on from the offer shown after you publish or start an import, at the top of your notifications, or under Notifications in your settings, where you can also choose which kinds arrive this way.

Changed

• Free accounts get five times as many AI tokens each month. The allowance is now 50,000 tokens, up from 10,000: enough to import or build about three arguments a month, with room to try one again.

• The respond form is easier to read. Its question shows the claim and whether you are saying it is true or false in colour, and its buttons are laid out as in every other claim form, with the main one alone at the right.

• A response's cards say what they answer. Each card answering a claim or a step opens with a label saying which, and its edge takes the same colour. The line under each card about whether its reasons were backed up is gone.

• A claim's origin mark is now shown for each place it appears. A claim placed twice, quoted from the source text at only one of those places, is original there and an enthymeme at the other. An original claim no longer shows a letter; its Source button says it is original.

• A premise's responses and menu stay on the first line of its title, and a long title now wraps instead of being cut off.

• Menu entries that delete something are drawn in red, and the premise's "Delete premise" is now "Delete". On an argument card, "Archive Permanently" is now "Archive", in amber, and "Delete Permanently" is "Delete", in red.

• An argument with no responses says "No published responses yet."

• Choosing a premise from the contents now closes them, so the premise lands where you expect.

• A source text no longer says the argument represents it faithfully. That line is gone from the Source text section and from an argument you export.

• The link to an argument's original post now reads "View on X", and an argument imported from a post reads "Imported from X".

• The report form now says what it is for, and can block the author at the same time. It opens with a short explanation and links to the Terms and Conditions and the Community Guidelines. Tick Also block this author to report and block in one go. The separate Block author button has left claim details and Argument details: use that option, or Block author in the argument's menu.

• An import's page now shows what the import was at the top. Its status, when it started and ended, and a link to its argument now head the page, above how the import went.

• Cancelling or restarting an import from your task list now asks first. The menu's Cancel and Restart ask the same question the import's own buttons do, and nothing happens if you say no.

Fixed

• Selecting a claim no longer moves its card. It used to shrink and shift slightly to the right as it was selected.

• Saving your settings no longer refuses a username you already have. A name that would not be allowed for a new account is accepted when it is the one your account already uses, and the settings page no longer shows an error under it.

• You can add a claim to a premise straight after adding the premise. Adding a claim to a premise you had only just added, before it had finished saving, used to fail with an error, and trying again from the same window failed too. The Add premise button now shows that it is working until the premise is ready, and the premise appears once it is, so everything on it works from the moment you see it. Renaming it, making it the conclusion and deleting it straight away work too.

• Adding a claim to a claim you have just negated no longer fails when you save it before the negation has finished saving.

• An import that failed before it began now says why, and can be retried. Its page used to say only that there was nothing to show. One that has not begun yet can be cancelled from its page.

• The reason given for a failed import no longer quotes an internal error. The reason at the top of its page, in each step's details, in your task list and in its notification now says why only when the reason was written for you, such as text the AI provider declined or claims that could not be connected to a conclusion; otherwise it says that something went wrong while building the argument.

• A new argument no longer changes the citation of an earlier one made from the same source text. Arguments you make from exactly the same text share that text, and its citation with it. Making a second argument from it with a different citation used to replace the first argument's citation without saying so, and once the text was published it stopped the new argument being made at all. Now the first citation stays, the new argument is made with it, and while you make the new argument you are told which citation the text already has, before you create anything.