# Kyurations > Kyurations is a curation tool for people who collect things worth keeping. You save a link, a note or a file from anywhere on the web, and it arrives as a card that looks like where it came from, with its cover, author, tags and a short gist already filled in. Lists form themselves out of what you save, you rearrange them until they feel right, and you publish the ones you want seen. It is free, your collection is yours, and you can export it at any time. Written by Krishna Kumar Soni, Founder. Builds Kyurations. Writes about curation, taste, and why most saved links never get opened again. The full text of the feature guide (10 chapters) and every editorial page (31). The index with links only is at https://www.kyurations.in/llms.txt. --- # The guide What the product does, chapter by chapter. This is the authoritative source: where this disagrees with a comparison page, this is right. ## Saving things Source: https://www.kyurations.in/guide/saving Updated: 2026-08-05 You save to Kyurations from wherever you found the thing. There is a Chrome extension for one-click saving from any page, an Add item box inside the app that takes a link, a note or a file, a bulk import that reads a bookmarks file or an export from Raindrop, Pocket, Instapaper and the rest, a Readwise import that brings highlights with it, and connections to X and Instagram that pull in what you bookmark and what people send you. Anything with a URL can also be saved by an AI assistant over MCP. Nothing asks you to choose a folder first. ### The extension The Chrome extension is the fastest way in, and the one most people end up using for almost everything. You are on a page worth keeping, you click the extension, and it saves. No tab switch, no copy and paste, no decision about where it goes. It reads the page you are on rather than just its address, so a save from the extension arrives with the same cover, author, tags and summary as a save made anywhere else. You connect it once by clicking Connect, which hands it a key tied to your account. It never asks for your password. Worth knowing: The extension is also how transcripts get fetched. YouTube will not hand a transcript to a server, but it will hand one to a browser, so the extension quietly does that job when a video you saved has captions. ### A link, a note, or a file Inside the app, Add item opens a box with three tabs. They cover the three things a person actually has: something they found, something they thought, and something on their machine. [Screenshot: The Add item dialog, open on the Link tab with an empty URL field] - Link takes any http or https address. It reads the page first and shows you what it found, so you are approving a save rather than making one blind. - Note has no URL and no title field. The body is the note. This is for the thought that has no link attached, which is most of the good ones. It is a proper writing surface: headings, lists, quotes, colour, links, images you paste or drop straight in, and any font on Google Fonts at any size. - File takes an upload and treats it as a first-class item, with metadata pulled from the format rather than from a web page. - Add item: a - New list: n ### Bringing in what you already have There are two bulk imports, and they do different jobs. Readwise walks your library, turns every source into an item and every highlight into an annotation on that item, so your marked-up passages arrive attached to the thing they came from rather than as a separate pile of quotes. Running it twice is safe. An already-imported highlight is a no-op, so importing again after a month of reading only adds the new ones. The token you paste is used for that one request and never stored. The second import reads a file, and it is the one for everything else: a bookmarks file from Chrome, Firefox, Safari, Edge or Arc, or an export from Raindrop, Pocket, Instapaper, Pinboard, Omnivore or Notion. Drop it into Settings. The file is read in your browser rather than uploaded, and what it says about each link, the title, the note and the tags, is kept instead of being thrown away and re-fetched. Your folders come across as DRAFT lists, never signed ones. That distinction is the whole design: a signed list is one you stamped as yours, and an archive you imported wholesale has not been chosen by anybody yet. You sign the ones that turn out to be real collections and leave the rest as drafts. Re-importing the same file changes nothing, because a link you already have is not saved twice. What it does not do: A large import lands immediately but arrives thin: every link is saved with the title from the file, then its cover, summary and tags fill in afterwards, a batch at a time. Settings shows how many are still waiting and lets you pick the work up later. ENEX from Evernote is not read; export those as a URL list instead. ### X and Instagram Connect X in Settings and your bookmarks come across. Turn on auto-sync and it keeps checking, at most once every fifty-five minutes, picking up where it stopped rather than starting over. A thread arrives as one item with the whole thread in it, not as one item per post. Instagram works the other way around. You connect your account, and anything you share to Kyurations in a DM gets saved. Posts and reels both, with the reel playable inline once it lands. [Screenshot: The Settings page, showing the Connected apps card with X connected] ### From an AI assistant Kyurations speaks MCP, so Claude, Cursor and other agents can save on your behalf. Hand one a URL and it goes through exactly the same path as a save you made yourself: extracted, summarised, tagged, deduped. Agents can also write notes with no link attached. This is covered properly in Kyuri and agents. The point here is that it is a real save path, not a side door with different rules. ### Saving the same thing twice You cannot end up with two copies of one page. A save is keyed on you and the URL, so saving something you already have finds the existing item instead of making another. If you saved it from the extension in March and again from your phone in August, there is still one card, and it is the one you have already annotated. Notes are the exception, because two notes with the same words are still two thoughts. ### Questions **Do I have to pick a list when I save something?** No. Everything lands in your Bower, which is the single surface holding everything you have saved. Lists are formed later, either by you or automatically out of what accumulates. Choosing a folder at save time is the step that makes people stop saving. **Can I save from my phone?** Yes. The web app works on a phone and can be installed to your home screen. Instagram shares also arrive from your phone, since sharing a reel to Kyurations in a DM saves it. **What happens if the page I saved goes offline later?** The cover image is copied to Kyurations at save time rather than hot-linked, and for articles the text itself is captured, so a card does not go blank when the source does. The original link stays on the card either way. ## What happens on save Source: https://www.kyurations.in/guide/enrichment Updated: 2026-08-04 When you save something, Kyurations reads it and fills in what a bookmark normally leaves blank: a real title, a short summary in plain language, a set of tags, the author or channel, and a cover image copied to Kyurations rather than hot-linked. Videos get their transcript where one is available, articles get their full text, and playlists arrive as a single item with the whole syllabus inside. You can edit any of it, and nothing waits on you to do so. ### What gets filled in A saved link is not much on its own. A URL and whatever the page happened to put in its title tag is enough to find something again and not enough to recognise it six months later. So every save is read before it lands. - A title that says what the thing is, rather than whatever the site called its tab. - A summary of a few sentences, written to be read rather than to be a preview of the first paragraph. - Tags, drawn to match the tags already in your library rather than invented fresh each time. - The author, channel or publication, so a card credits the person who made it. - A cover image, copied into Kyurations as WebP at save time. [Screenshot: A saved YouTube course, showing an automatic summary, seven tags, and the full episode list] ### Cards that look like where they came from A tweet does not look like a YouTube video, and neither looks like a PDF. Kyurations keeps that difference instead of flattening everything into one rectangle with a favicon on it. A card carries the shape of its platform: a post shows its author and its text, a video shows its still and its duration, a reel shows as a reel. The reason is recognition. You do not remember a saved thing by its title. You remember it as the video with the whiteboard, or the thread by that person, and a card that keeps its native shape lets you find it the way you actually remember it. [Screenshot: A saved post from X with an automatic summary and three tags] ### Pages without a picture Plenty of good pages have no cover image. Those do not get a grey box. They become a text-led card with a large favicon tile, named by the publication rather than by the word Web, which tells you nothing. ### Video, audio and long text A video you saved is worth more with its words attached. Where a video has captions, the transcript is fetched and stored with the item, so it can be read, searched and asked about. Reels with no captions at all are transcribed by listening to them. Articles have their full text captured at save time, which is what makes the reader work and what keeps a card useful after the source has gone. Each one is scored as complete or partial, so you know whether you are reading the whole thing. A YouTube playlist is saved as one course rather than as forty separate videos, with its syllabus intact. A thread on X is saved as one item containing the thread. ### Fixing what it got wrong Every field is editable, in place, on the card. Titles, summaries, tags and notes are text you can click into. Nothing opens a form. If a page changed after you saved it, or the save caught it mid-load, refresh from source re-reads the page and rebuilds what it found. - Edit the notes: e - Edit tags: t - Refresh from source: r What it does not do: Instagram profile pages are the one thing that cannot be read from a server. Everything else on the list of platforms is extracted without a browser in the loop. ### Questions **Which AI writes the summaries?** Summaries and tags are written by Claude. Search across your library uses embeddings from OpenAI. Neither is trained on your collection. **Can I turn the automatic summary off?** There is no switch for it, but nothing is permanent. Every field the save filled in is editable in place, and a summary you rewrite stays rewritten. **Why do my tags look like tags I already use?** Because they are matched against your existing tags before new ones are invented. A library where the same idea is filed under three near-identical tags is a library you cannot filter, so tagging is reconciled against what you already have. ## Your Bower Source: https://www.kyurations.in/guide/bower Updated: 2026-08-04 The Bower is the single surface holding everything you have saved. It has no folders. You change how it looks with three layouts, narrow it with tags and facets, and act on many things at once by selecting them. It is named after the bowerbird, which collects objects and arranges them for the pleasure of looking at the arrangement. ### Why there are no folders Filing is the step that kills collections. Deciding where something goes is a small tax charged at the exact moment you are least willing to pay it, and a tool that charges it gets used for a month. So everything you save lands in one place, and arrangement happens later, when you are in the mood to arrange rather than in the middle of reading something else. Lists are the arrangement. The Bower is everything. [Screenshot: The Bower in grid view, showing saved posts, videos and notes as cards of different shapes] ### Three ways to look at it The same collection, three layouts. Which one is right depends on what you are doing, so the toggle is one keypress and the choice sticks. | Layout | Shows | Good for | | --- | --- | --- | | Zen | Summary text, no card chrome, no images | Reading through what you have, when the pictures are a distraction | | Grid | Platform-true cards in a masonry wall | Recognising things, which is how you actually find them | | List | Dense metadata rows | Scanning a lot at once, and bulk work | [Screenshot: The Bower in zen view: columns of summary text with no images or card borders] - Switch view: v - Shuffle the Bower: s - Refresh: r Worth knowing: Shuffle is not a gimmick. A wall in the order you saved things is a wall where the last month is all you ever see. Shuffling surfaces the things you saved in March and have not thought about since. ### Narrowing it down Tags run across the top as chips. Click one to narrow to it. Alongside them are three facets that answer the questions tags cannot. - Sort by, for changing the order rather than the contents. - Platform type, for when you want the things from one place: everything from YouTube, everything from X. - Media type, for when you want one kind of thing: every video, every note, every file. - Clear all filters: Shift+f ### Working the grid from the keyboard The grid has a cursor. You can move through it, open things, and act on them without reaching for the mouse. This is the part that turns a browse into a session. - Next card: j or ArrowDown - Previous card: k or ArrowUp - Card to the right: ArrowRight - Card to the left: ArrowLeft - Open: Enter - Open at source: o ### Doing something to many things at once Select cards and a bar appears with what you can do to all of them. Moving forty things into a new list is one action, not forty. - Select this card: x - Select everything visible: Shift+a - Move to a list: m - New list from selection: n - Open all at source: o - Delete selected: Backspace or Delete ### Undoing a delete Deletes are optimistic, which means the card goes immediately rather than after a confirmation dialog and a spinner. That is only reasonable if the mistake is cheap to fix, so the last delete is undoable. - Undo the last delete: Mod+z What it does not do: Undo covers the last delete, not an unlimited history. A deletion made from an AI assistant over MCP is a soft delete to the trash and can be restored from there instead. ### Questions **Why is it called the Bower?** A male bowerbird builds a structure out of collected objects, arranges them by colour and size, then steps back and moves things until it looks right. The bower has no function beyond being looked at. That is the test for whether a collection has been curated, and it is what the surface is named after. **Does the Bower have a search?** Command-K opens a palette that searches and runs commands, and Kyuri answers questions across everything you have saved in plain language. For narrowing rather than asking, the tag chips and the three facets are usually faster than typing. **How many things can it hold?** There is no limit set on it. The grid loads as you scroll rather than all at once, so a large Bower opens as fast as a small one. ## Reading and marking up Source: https://www.kyurations.in/guide/reading Updated: 2026-08-04 What you saved opens inside Kyurations rather than bouncing you to the original site. Articles open in a reader, PDFs open with real text you can select, videos carry their transcript, and reels play as reels. You can highlight passages, clip a region of a page, and write a note against any of it, and those marks stay attached to the exact words they were made on. ### The reader Open a saved article and you get the text, set to be read, with the summary and your notes in a panel beside it. The original is one click away and mostly you will not want it. [Screenshot: A saved article open in the reader, with the article text on the left and a notes panel on the right] - Hide or show the side panel: z - Full screen: f - Next item: ArrowRight - Previous item: ArrowLeft ### Highlights Select text and highlight it. The passage stays marked, and it is listed with the item so you can come back to the four sentences that mattered rather than to the whole article. [Screenshot: A paragraph of an article with a sentence highlighted, and the highlight listed in the side panel] Highlights are anchored to the words rather than to a character offset, using the same matching a browser uses for its own text fragments. That is what lets a highlight survive the page being re-fetched and the text shifting underneath it. Worth knowing: Highlighting works on a phone. A long press and the handles are enough; there is no separate mode to enter. ### PDFs A saved PDF opens in the same place as everything else, with selectable text rather than as a picture of a page. That means it can be highlighted like an article can. Zoom is a set of measured steps rather than a text size, because a PDF has a fixed layout and growing the type would break it. ### Clipping part of a page Some of what is worth keeping is not text. A chart, a diagram, a screenshot inside an article. You can clip a region of the page and keep it as a mark on the item, alongside the highlights, with a note attached if it needs one. ### Video, reels and playlists A saved video plays in place. Where a transcript was found, it sits beside the video, which makes a forty-minute talk searchable and skimmable. A set of saved reels plays as reels: full height, one after another, the way the format is meant to be watched. That works on a list somebody shared with you as well as on your own. A saved course keeps its episode list, so you can see all thirteen videos and where you are in them without leaving. [Screenshot: A saved web page open in the detail view, showing the page image, a note and three tags] ### Notes on anything Every item has a note field, editable in place. It is the most useful thing on the card and the most skipped: one line on why this is here does more work six months later than any tag. - Edit the notes: e - Edit tags: t - Copy link: c - Move to a list: m ### Questions **Do my highlights survive if the article changes?** They are matched on the words themselves rather than on a position in the document, using the browser's own text-fragment algorithm. Small edits to the page around a highlight do not lose it. A passage deleted outright cannot be found, and that highlight stays listed with the item. **Can I read something that has gone offline?** Usually. Article text is captured at save time and stored with the item, so the reader still has something to show after the source has gone. Each capture is scored as complete or partial so you can tell what you are looking at. **Does it read paywalled articles?** It captures what the page gives it. If a site shows the first two paragraphs to everyone, that is what is stored. Kyurations does not get past paywalls, and does not pretend to. ## Lists Source: https://www.kyurations.in/guide/lists Updated: 2026-08-04 A list is a small set of things, in an order you chose, with a name that says something. Lists in Kyurations form two ways: you make one, or the product notices that a group of things in your Bower belong together, names it, and offers it to you as a draft. A draft is a suggestion until you sign it, at which point it is yours. ### What a list is for A list is not a folder and it is not a tag. A folder is where a thing lives, and a tag is what a thing is about. A list is an argument: these nine things, in this order, under this name. That is why the same item can sit in several lists without being copied, and why a list called Reading is not a collection while a list called Things that changed how I write is. [Screenshot: The Lists view, showing seven list covers with names, descriptions and item counts] ### Lists that form themselves Most people never get around to arranging what they save. So the product does a first pass. When enough related things accumulate, they are grouped, given a name, and offered as a list. Every save is also read against the lists you already have, and offered to the ones it genuinely belongs in. Make a list called Design Libraries for AI SaaS and it fills itself with design libraries, not with everything that happens to mention AI. Nothing to press: it happens as you save, and anything filed automatically stays marked as a suggestion until you touch it. When a suggestion is wrong, flick it off. That is not just a delete: it is recorded as a standing instruction, so the same kind of thing stops being offered to that list, and to lists you make later. ### Signing a draft A draft is rendered faintly, held back from the others, because it is a hypothesis and not yet your work. It carries one affordance: Sign to claim. Signing is the moment a machine's guess becomes a person's collection. Before it, nobody has vouched for the grouping. After it, you have, and the list behaves like any list you made yourself. A draft you do not want is dismissed and does not come back. - Sign a draft: Shift+s ### Building one by hand New list, name it, add things. Inside a list, both Add buttons open a picker of what you have already saved, because a list is made out of your Bower rather than out of the open web. Capture lives on the Bower; arrangement lives here. The picker doubles as a create surface, so if the list you want does not exist yet you can make it without backing out first. - New list: n - Add items: a - Rename: e - Duplicate: d - Delete list: Backspace or Delete ### Adding to the right list When something is proposed for a list, it is judged against what that list is actually about rather than against every item that happens to be in it. A list of nine essays on typography with one stray link about coffee is still a list about typography, and admission is decided on the nine. Worth knowing: An item in two lists is one item. Edit its note in one place and the other list shows the same note, because there is no copy. ### What a list carries - A name and a description, both editable in place. - A cover, taken from what is inside it, and a colour derived from that cover. - Its own tags, which are the tags of the things in it. - A visibility: private, public, or reachable with a link and a password. - Optionally a price, and optionally a newsletter. ### Questions **Does putting something in a list take it out of my Bower?** No. The Bower holds everything, always. A list is a view onto some of it, and the same item can be in as many lists as it belongs in. **What happens to a draft I ignore?** It sits there, faint, until you sign it or dismiss it. It is not published, not counted as yours, and not shown to anybody else. **Can I reorder a list?** Yes, and the order is part of the argument. The same twelve links in a different order say a different thing. ## Sharing and publishing Source: https://www.kyurations.in/guide/sharing Updated: 2026-08-04 Any list can be published as a page anybody can open, sent privately to named people, or put behind a link and a password. A published collection gets its own address, a cover, likes and view counts, and can appear on Discover and on your profile. It can also run as a newsletter, mailing its newest items to subscribers every one, two or four weeks. ### One switch, three states Sharing is a single control rather than a scatter of settings. A list is private, public, or open to anyone holding both the link and a password. Turning one on turns the others off, and the modal says what changed. [Screenshot: The share dialog with the public link switched on, showing the link, a Sell toggle, a Subscription toggle and a frequency menu] ### Publishing to everyone A public list becomes a real page: its own address, its cover, its description, and every item laid out the way it is laid out for you. Reels play, videos play, and a reader can open any item on its own. It carries a like count and a view count, and a Curate your own invitation for the visitor who wants one of their own. It is also indexable, which is how somebody who has never heard of you finds a collection you made. [Screenshot: A published collection page titled AI Engineering Fundamentals with sixteen items and subscribe, like and view counts] ### Sending it to particular people Not everything worth sharing is worth publishing. Share Privately sends a list to named people by handle or by email. They get it in their Shared tab, wearing the same cover it wears for you, and the list stays off the public web. [Screenshot: The share dialog on the Share Privately tab, showing two people with view access and a collapsed link-and-password section] For a reader who has no account and should not need one, the third state is a link with a password on it. Anyone with both can open the page. Anyone with only the link cannot. ### Running a list as a newsletter A public list can collect subscribers. Turn Subscription on and a Subscribe button appears on the page; from then on, new things you add are mailed out on a schedule. A subscriber is someone with a Kyurations account who subscribed from the list's own page, and the issue goes to the verified email already on that account. There is no box where a stranger types an address, and no way to import one. - The cadence is every week, every two weeks, or every four. Two weeks is the default. - Each issue carries up to five items, written up as an issue rather than listed as links. - Everything added since the last issue leads. If fewer than five things are new, the issue is topped up with the most recent older members so it still reads like an issue. - Send yourself a sample first. It is the same email your subscribers will get. Worth knowing: A list running a newsletter is marked Live, and a Live list is the one place per-list numbers are shown: subscribers, views, likes, item count and when the next issue goes out. There is a page showing what an issue actually looks like at kyurations.in/newsletters. ### Your profile, and Discover Your handle is a page. Everything you have published sits on it, at an address you can put in a bio. Published collections also appear on Discover, which is the front door for people browsing rather than searching, and on subject pages grouping collections by what they are about. Every published page carries the structured description search engines and AI assistants read, and a preview card built for it, so a link posted anywhere arrives looking like something rather than like a URL. ### Taking it back Switch sharing off and the page stops resolving. The share link is not recycled, the list is intact, and nothing you wrote is lost. A password link works the same way. - Share: s - Like this list: l - Subscribe: b ### Questions **Do people need an account to read a collection I published?** No. A public collection opens for anybody, signed in or not. Accounts are only needed to save, to curate, and to buy. **What do subscribers actually receive?** An email carrying up to five items, led by everything added since the last issue and topped up with recent older members if fewer than five are new, on the cadence you set. If you added nothing at all, nothing goes out. **Can I see who subscribed?** You see the count, and on a Live list you see it alongside views, likes and the next issue date. Kyurations does not hand you a list of subscriber email addresses. ## Selling a collection Source: https://www.kyurations.in/guide/selling Updated: 2026-08-04 Any public list can carry a price. Buyers read the first few items free, then pay once for permanent access to the rest, including anything you add later. Kyurations keeps 30 percent and you keep 70. Payouts go out weekly to a UPI ID or a bank account, and buyers can ask for a refund within seven days. ### How a paid collection reads A paid list is not a locked door. The opening items are readable by anybody, in full, exactly as they would be on a free collection. The page then says plainly how much more there is and what it costs. [Screenshot: A collection for sale, showing two free items and a buy panel offering the rest for four dollars] The sample is the argument. Somebody who has read two of your picks and wants the other twelve is a much better buyer than somebody staring at a blurred grid. ### What a buyer gets - The whole collection, permanently, for one payment. There is no subscription. - Everything you add to that list afterwards, at no further cost. A collection that keeps growing is worth more than a collection frozen at the moment of sale. - Access tied to their account rather than to a link, so it survives a new device. - Seven days to ask for a refund if the collection was not what the page described. Worth knowing: A purchase is a grant, not a copy. The buyer is reading your list, so when you improve it, their copy improves. Nothing is duplicated into their library. ### Setting a price 1. Publish the list: Selling runs on top of a public link, so the list has to be public first. 2. Turn on Sell: It sits in the same share dialog as the public link, under Quick settings. 3. Set the price and the free sample: Choose what it costs and how many items stay open to everyone. 4. Add a payout destination: In Settings, under Payouts. Without one the money is still owed to you, it just has nowhere to go. ### The split, and getting paid You keep 70 percent of each sale. Kyurations keeps 30, which covers payment processing, the merchant-of-record obligations and the hosting. Payouts run weekly, every Friday, to a UPI ID or a bank account. A destination counts as verified only once money has actually landed in it, so a freshly entered UPI ID reads as saved rather than as confirmed until the first payout clears. [Screenshot: The Settings page showing the Payouts card with a saved UPI destination and a weekly Friday schedule] The Earnings page shows what has sold, what is pending and what has been paid. It only appears for people who actually sell. ### Refunds and disputes A buyer who writes in within seven days, because the collection was not what the page said it was, gets their money back and loses access. Refunds are reconciled against the original payment, so a refunded sale stops counting toward your earnings rather than being paid out and clawed back later. ### Questions **What does it cost to sell?** Nothing up front. Kyurations takes 30 percent of a sale and you keep 70. If nothing sells, nothing is charged. **Can I sell a list and also run it as a newsletter?** Yes. They are separate switches on the same public list. A paid collection with a newsletter keeps mailing its buyers as it grows. **What currency do buyers pay in, and what do I get paid in?** Prices are set in dollars. Payouts to a UPI destination are made in rupees. The Earnings page shows both what was charged and what is owed to you. **Do buyers keep access if I delete the list?** Deleting a collection you have sold takes away what somebody paid for, so it is not something to do casually. The commercial terms, including what buyers are owed, are set out on the selling page. ## Kyuri and AI tools Source: https://www.kyurations.in/guide/kyuri Updated: 2026-08-04 Kyuri is an assistant that reads your own collection. Ask it a question in plain language and it searches everything you have saved, answers, and cites the specific items it used so you can check the claim. Kyurations also runs an MCP server, so Claude, Cursor and other agents can search, read, save, tag, organise and publish on your behalf from wherever you already work. ### Asking your own collection The useful thing about a collection you built is that you already vetted it. Kyuri answers from that, and only from that. It is not a general chatbot with your bookmarks bolted on. [Screenshot: A Kyuri answer summarising several saved items, with a collapsed references section showing sixteen sources] Every answer carries its references. Open them and you get the actual items, so a claim you doubt takes one click to check. If nothing in your collection answers the question, it says so rather than filling the gap. ### Narrowing what it reads The composer carries a scope chip. Leave it on all lists and Kyuri reads everything; point it at one list and it reads only that. Asking a question of a single collection is how you get an answer about that subject rather than an answer that drags in everything adjacent. - Ask Kyuri: / - Go to Kyuri: g k - New chat: Mod+Shift+o - Focus the composer: / ### Taking a conversation elsewhere A chat can be handed off to ChatGPT or Claude, carrying what you were talking about with it, for when the next step needs a general model rather than one confined to your library. - Hand off this chat: Mod+Shift+e ### Connecting your own AI tools Kyurations speaks the Model Context Protocol. Generate a token in Settings, paste the endpoint and the token into your client, and your library becomes something the assistant you already use can work with. Reading tools, for asking: - Search the library, semantically and by keyword, with a link on every result. - Filter by platform, type, author, tag, date or a named list. - Count and break things down: how many videos, what you save most about, how much sits where. - Read an item in full, including its transcript, with its source, author and tags. - See your lists, with sizes and public links. Writing tools, for curating: - Create a list and fill it, adding and removing items as it takes shape. - Save a new link, extracted and deduped exactly as if you had saved it yourself. - Write a note with no link attached. - Edit what is saved, changing only the field named. - Tag and retag across items. - Publish a list and get its link back, or take it private again. - Delete to the trash, and restore from it. Marked destructive, so a good client asks first. Worth knowing: Tokens are per-client and revocable. Connected apps in Settings lists everything holding one, with the date it was last used, and revoking is immediate. ### The AI Kyurator There is a second thing in here, which goes the other way: rather than reading what you saved, it goes and finds things it thinks match your taste, and proposes them. Not open yet: The AI Kyurator is not open yet. It runs on the maker's own account while the taste model is tuned, because a recommender that has not earned its judgement is just noise with a nice interface. Everything else on this page is available to everyone today. ### Questions **Does Kyuri make things up?** It answers from your collection and cites the items it used. Where your collection has nothing to say on a question, the honest answer is that it has nothing, and that is what you get. **Is my collection used to train a model?** No. Summaries are generated by Claude and search runs on embeddings from OpenAI, both as ordinary API calls. Your library is not training data. **Which MCP clients work?** Any client that speaks MCP. Claude, Cursor and the rest connect the same way: an endpoint and a token, both generated in Settings. ## Your account and your data Source: https://www.kyurations.in/guide/account Updated: 2026-08-04 Settings holds your handle, photo and bio, the apps you have connected, your payout destination and your keyboard preferences. Everything you have saved can be exported at any time, per list or as a whole account, in four formats. There is no lock-in and no export fee. ### Your handle and profile Your handle is your address. It is where your published collections live, and it is what you put in a bio somewhere else. Three to twenty characters, lowercase letters, numbers and underscores. You can change it up to three times in any rolling three months. The limit exists because a handle is a public address other people have linked to, and a name that moves every week is not an address. A photo and a one-line bio sit beside it. Both show on your profile and on every collection you publish. [Screenshot: The Settings page showing profile, account, payouts, MCP set-up and connected apps] ### Connected apps - X, for pulling in your bookmarks, with auto-sync as an option. - Instagram, so anything you share to Kyurations in a DM gets saved. - The Chrome extension, for one-click saving from any page. - Any MCP client holding a token, each listed with when it was last used. Every one of them disconnects from the same place, and disconnecting takes effect immediately. ### Keyboard preferences Single-key shortcuts can be turned off. This is for people who dictate: speech input types letters into whatever has focus, and a bare `d` that deletes something is a bad interaction with a microphone. Turning them off leaves the modifier shortcuts working. ### Taking everything out Export gives you what you saved, in a form something else can read. It works on one list or on the whole account. | Format | What you get | For | | --- | --- | --- | | Markdown | A zip of files, one per list, with notes and highlights | Obsidian, and anything else that reads plain text | | CSV | One row per item, with its fields as columns | A spreadsheet, or a script | | Bookmarks | A Netscape bookmark file | Importing into a browser | | JSON | The full structure, nothing dropped | Moving to another tool, or keeping a real backup | What it does not do: The bookmark format was invented in the nineties and has room for a name, a URL and a date. Notes, tags and highlights do not survive it. The other three formats carry everything, and the export dialog says so before you pick. ### What is stored, and what is not Your collection is yours. It is not sold, not used as training data, and not mined to build a profile of you. The privacy page sets out what is kept and why, in the same plain language as the rest of this. Some credentials are deliberately never stored at all. The token you paste to import from Readwise is used for that one request and dropped when it finishes. ### Telling us something is wrong There is a feedback item in the account menu. It writes to the person who builds this, not to a queue. The release notes are the other half of that conversation: everything that ships is listed there, dated. ### Questions **Is Kyurations free?** Yes. Saving, curating, publishing, asking Kyuri and exporting all cost nothing. The only money that changes hands is when somebody sells a collection, and then Kyurations takes a share of that sale. **Can I export and leave?** Yes, at any time, in four formats, with no fee and nothing withheld. A tool that makes leaving hard is a tool that has stopped trying to be worth staying for. **Does it work on a phone?** Yes. The web app is built for a phone as well as a desktop and can be installed to the home screen. Highlighting, reading and reels all work by touch. ## Keyboard shortcuts Source: https://www.kyurations.in/guide/shortcuts Updated: 2026-08-04 Kyurations is built to be worked from the keyboard. Command-K opens a palette that searches and runs commands, a question mark shows the same list you are reading now inside the app, and single letters act on whatever is focused. Bare keys never fire while you are typing, and they can be switched off entirely if you dictate. ### How the bindings work Shortcuts are scoped rather than global. The innermost thing that is active and has a handler wins, which is why r refreshes a focused item from its source but refreshes the whole library when nothing is focused. Neither one needs to know about the other. Bare letters do not fire while you are typing into a field. Modifier shortcuts do, because a palette you cannot open from a search box is a palette you reach for and miss. [Screenshot: The in-app keyboard shortcuts panel, listing general, go-to and item bindings in two columns] Worth knowing: Bindings show here as your own machine spells them: the Command glyph on a Mac, Ctrl elsewhere. ### General Everywhere, whatever you are looking at. - Search and run commands: Mod+k - Keyboard shortcuts: ? or Shift+/ - Ask Kyuri: / - Settings: Mod+, - Toggle sidebar: [ - Undo the last delete: Mod+z - Close, go back, or clear: Escape - Cancel: Escape - Save: Mod+Enter ### Go to Two keys: press g, then the letter. Chords rather than modifiers, so nothing collides with the browser. - Go to Bower: g b - Go to Lists: g l - Go to Shared: g s - Go to Kyuri: g k - Go to the Feature guide: g f - Go to Release notes: g r ### Library On the Bower, Lists and Shared. - Add item: a - New list: n - Switch view: v - Refresh: r - Shuffle the Bower: s - Clear all filters: Shift+f ### Moving around The grid has a cursor. These move it. - Next card: j or ArrowDown - Previous card: k or ArrowUp - Card to the right: ArrowRight - Card to the left: ArrowLeft - First card: Home - Last card: End - Select this card: x - Select everything visible: Shift+a ### Item When a card is focused, or its detail view is open. - Open: Enter - Open at source: o - Edit the notes: e - Edit tags: t - Move to a list: m - Duplicate: d - Copy link: c - Refresh from source: r - Remove from this list: - - Delete: Backspace or Delete - Full screen: f - Hide or show the side panel: z - Next item: ArrowRight - Previous item: ArrowLeft ### Reels In the full-screen reels player, which sits above everything else while it is open. - Next reel: j or ArrowDown - Previous reel: k or ArrowUp - Mute or unmute: m - Close the player: Escape ### List When a list card is focused, or you are inside one. - Open: Enter - Add items: a - Share: s - Rename: e - Duplicate: d - Analytics: i - Sign a draft: Shift+s - Delete list: Backspace or Delete ### Selection Once one or more cards are selected. - Move to a list: m - New list from selection: n - Open all at source: o - Delete selected: Backspace or Delete ### Kyuri In the chat. - New chat: Mod+Shift+o - Hand off this chat: Mod+Shift+e - Focus the composer: / - Send: Enter - New line: Shift+Enter ### Public list For a visitor reading a collection somebody published. - Like this list: l - Subscribe: b - Read or gallery: v ### Questions **How do I see shortcuts inside the app?** Press the question mark key. The panel it opens is built from the same registry as this page, so the two cannot disagree. **Can I turn single-key shortcuts off?** Yes, in Settings. It is there for people who use speech input, where dictation types bare letters into whatever has focus. Modifier shortcuts keep working. **Can I rebind a key?** Not today. The bindings are fixed, and the only preference is whether single-key ones fire at all. --- ## Pocket alternatives, one year after the shutdown Source: https://www.kyurations.in/alternatives/pocket Updated: 2026-08-05 Mozilla shut Pocket down on 8 July 2025 and deleted all remaining user data on 8 October 2025, so there is nothing left to migrate from. The best replacement depends on what you used Pocket for. Raindrop.io is the closest free swap for saving and tagging links, Readwise Reader is the strongest choice if you read and highlight seriously, Instapaper is the lightest reader, and Kyurations is the one to pick if the point was never storage but building a collection you would show someone. Mozilla announced the shutdown on 22 May 2025, closed the service on 8 July, and gave a three-month export window that ended on 8 October 2025. Anyone who missed that window has no archive to move. If you have an export file sitting on a drive, Raindrop, Readwise Reader, Karakeep and Kyurations will all read it directly. Pocket did two jobs at once: it was a queue of things to read later, and it was a pile of things you liked. Most replacements only do one of those well, which is why picking the wrong one feels wrong for months. ### Raindrop.io (Free, with a paid tier around $3 a month) The closest free swap for saving, tagging and finding links again. Best for: Someone who wants Pocket back, with folders. - Strength: Unlimited bookmarks on the free tier, which no other established tool matches - Strength: Nested collections, tags and a fast search - Strength: Browser extensions and apps on every platform - Limit: The reading experience is thinner than Pocket's was - Limit: Collections look like a file manager, so a shared collection reads as a folder rather than as something you made ### Readwise Reader (Around $10 a month) The best of these if you actually read what you save and mark it up. Best for: Someone whose saved links are research, not a wishlist. - Strength: Highlighting, notes and spaced review that surfaces old highlights - Strength: Handles articles, PDFs, email newsletters, YouTube and Twitter threads in one queue - Strength: Exports cleanly into Obsidian, Notion and Roam - Limit: No free tier worth planning around - Limit: Built for a private workflow, so there is little to show anyone ### Kyurations (Free) The one to pick if the pile was always meant to become something other people could see. Best for: Someone who saves with taste and wants the result to look like it, not like a folder. - Strength: Saved links arrive as cards that look like where they came from, with cover, author, tags and a short gist already filled in - Strength: Articles and PDFs read in the app, and it picks up where you stopped - Strength: Lists form themselves out of what you save, and you rearrange them until they feel right - Strength: Any list can be published as a page worth sending to someone - Strength: Ask Kyuri questions across everything you have saved, and every answer cites the items it came from - Strength: Export the whole collection at any time - Limit: Newer than the rest of this list, so it installs from the browser rather than from an app store - Limit: Reading happens in the app but not offline, and nothing keeps an unread count for you ### Instapaper (Free, with a paid tier around $6 a month) The lightest reader here, and the least in your way. Best for: Someone who wants text, on a phone, on a train. - Strength: The cleanest article view of any of these - Strength: Genuinely good offline reading - Strength: Has outlived most of its competitors, including Pocket - Limit: Organisation stops at folders and highlights - Limit: Not built for links that are videos, repos or products ### Karakeep (Free, self-hosted) The self-hosted option, if you would rather own the server. Best for: Someone who runs their own things and wants no vendor to trust. - Strength: Open source, so a shutdown announcement can never apply to you - Strength: Full-text search over saved pages, with automatic tagging - Strength: Docker install, active development - Limit: You are the operations team - Limit: Sharing is a link to your own server, which is not the same as publishing ### Pinboard (Around $22 a year) Plain text bookmarking that has not changed in fifteen years, on purpose. Best for: Someone who wants a URL, a tag, and no opinions. - Strength: Fast, tiny, and completely uninterested in engagement - Strength: Archiving tier keeps a copy of every page you save - Limit: No covers, no previews, no reading view - Limit: One person maintains it, which is both the appeal and the risk ### How to choose without trying six of them Ask what you did with Pocket on the day you liked it most. If the answer is that you read a long article on a phone, take Instapaper or Readwise Reader. If the answer is that you found a link you had saved eight months earlier, take Raindrop.io or Pinboard. If the answer is that you sent someone a list of the good stuff, take Kyurations. Most people who bounce between tools are answering the wrong question. They compare feature lists when the real difference is what the tool assumes you are doing. ### If you still have a Pocket export The export is an HTML file of links with tags and timestamps. Raindrop, Readwise Reader, Karakeep and Kyurations all read it directly. In Kyurations the tags come across with the links and the whole file lands in one go, though it is still worth asking how much of an archive you kept because it was good rather than because saving was free. If you do not have the file, you do not have the data. Mozilla deleted it after 8 October 2025 and there is no recovery path. ### Questions **Is Pocket really gone?** Yes. Mozilla closed the service on 8 July 2025 and permanently deleted all remaining user data after 8 October 2025. The apps and extensions no longer work. **What is the closest thing to Pocket now?** Raindrop.io, if you mean the saving and tagging. Instapaper, if you mean the reading. Neither is a full replacement, because Pocket was doing both. **Is there a free Pocket alternative?** Raindrop.io has the most generous free tier for plain bookmarking. Kyurations is free and adds automatic covers, tags and gists, plus publishing. Karakeep is free if you host it yourself. **Can I still recover my Pocket data?** Only if you downloaded the export before 8 October 2025. After that date Mozilla deleted it, and there is no way to request a copy. **Which Pocket alternative is best for sharing a reading list?** Kyurations, because publishing is the point of the product rather than a setting inside it. A list becomes a page with covers and your name on it in one tap. Raindrop can share a collection, but it arrives looking like a folder. --- ## Raindrop.io alternatives, and when to actually switch Source: https://www.kyurations.in/alternatives/raindrop-io Updated: 2026-08-05 Raindrop.io is the strongest free bookmark manager available, so most people looking for an alternative are not unhappy with it, they have outgrown one specific thing. If you want highlights and a real reading workflow, move to Readwise Reader. If you want your collections to be worth showing to other people, move to Kyurations. If you want to own the server, move to Karakeep. If you want less software rather than more, move to Pinboard. Raindrop absorbed most of Pocket's users in 2025 and has the best free tier in the category, so switching away from it is usually about fit rather than about faults. The three reasons people give: the reading experience is thin, collections look like folders when shared, and the paid tier gates the features they came for. ### Kyurations (Free) For when the collection is meant to be seen, not stored. Best for: Someone whose saved links are a portfolio of their taste. - Strength: Cards that look like their source, so a list of links looks like a magazine spread rather than a directory listing - Strength: Lists form themselves out of what you save, then you arrange them - Strength: Publish any list as a page, with your handle on it - Strength: Free, with export at any time - Limit: Fewer years behind it than Raindrop - Limit: Nested folders are not the model here, semantic lists are ### Readwise Reader (Around $10 a month) For when you want to read and mark up, not just file. Best for: Researchers, writers, and anyone whose notes matter more than their links. - Strength: Best-in-class highlighting and note capture - Strength: Newsletters, PDFs and video in the same queue - Strength: Deep export into note apps - Limit: No meaningful free tier - Limit: Private by design, nothing to publish ### Karakeep (Free, self-hosted) For when you want the software but not the vendor. Best for: Self-hosters. - Strength: Open source and actively developed - Strength: Automatic tagging and full-text search over archived pages - Limit: You run and back up the server - Limit: Publishing means exposing your own host ### Pinboard (Around $22 a year) For when the answer is less software. Best for: People who found Raindrop too visual. - Strength: Extremely fast - Strength: Optional archiving of every page you save - Limit: No previews or reading view - Limit: A one-person operation ### Where Raindrop is still the right answer If what you need is a large, searchable, well-tagged store of links across nested folders, and the free tier covers you, stay. Nothing here beats it on that job, and switching costs you a working system. The case for moving is narrower than the comparison sites suggest. It comes down to whether your links are an archive or an argument. Raindrop is built for archives. ### Questions **Is there a better free alternative to Raindrop.io?** For raw bookmark storage, no. Raindrop's free tier is the most generous in the category. Kyurations is free and better if you want the collection to be publishable and to look like something, but it is not trying to be a folder tree. **Can I move my Raindrop collections somewhere else?** Yes. Raindrop exports to CSV and HTML, which Readwise Reader, Karakeep, Pinboard and Kyurations all read. In Kyurations your collections arrive as draft lists you sign or discard, rather than as finished ones you never chose. **What is the difference between Raindrop.io and Kyurations?** Raindrop stores and organises links. Kyurations arranges them for display. Practically: Raindrop gives you folders and a list view, Kyurations gives you platform-shaped cards, self-forming lists, and a publish button that produces a page you would send to someone. --- ## Notion Web Clipper alternatives for saving links Source: https://www.kyurations.in/alternatives/notion-web-clipper Updated: 2026-07-27 The Notion Web Clipper saves a page title and a URL into a database row, and leaves you to fill in everything else by hand. If you want the cover, author, tags and a summary to arrive on their own, use Kyurations or Raindrop.io. If you want to read and highlight, use Readwise Reader, which also syncs highlights back into Notion. Stay with the clipper only if the saved link needs to live next to other Notion data. The clipper is a database row, which is exactly right when the link belongs inside a project and exactly wrong when it is the thing you are collecting. The usual breaking point is around two hundred saves, when the database is full of untitled rows nobody will ever revisit. ### Kyurations (Free) Saves the whole link, not just its address. Best for: Anyone tired of hand-filling database properties. - Strength: Cover, author, platform, tags and a short gist are fetched at save time - Strength: Notes and files sit alongside links, so a collection is not URL-only - Strength: Lists group themselves, and any of them publishes as a page - Limit: Not a database, so there are no formulas or relations - Limit: No Notion two-way sync ### Raindrop.io (Free, paid tier around $3 a month) The straight upgrade if you liked the database but hated the typing. Best for: People who want folders and speed. - Strength: Fast clipper with previews - Strength: Nested collections and tags - Limit: Shared collections look like folders ### Readwise Reader (Around $10 a month) The one that puts the highlights back into Notion for you. Best for: Notion users who mostly wanted the quotes. - Strength: Highlight sync into Notion - Strength: Handles PDFs and newsletters - Limit: Paid only ### Questions **Why does the Notion Web Clipper save so little?** It is a database row builder, not a link parser. It captures the page title and URL and leaves the remaining properties for you to fill in, because Notion has no opinion about what a saved link should contain. **What is the best free Notion Web Clipper alternative?** Kyurations and Raindrop.io are both free and both fetch the cover, author and tags automatically. Pick Kyurations if you want to publish the result, Raindrop if you want folders. **Can I keep using Notion and still save links elsewhere?** Yes, and most people should. Keep project material in Notion, keep the collection somewhere built for links, and export between them when you need to. --- ## Instapaper alternatives when reading is not the whole job Source: https://www.kyurations.in/alternatives/instapaper Updated: 2026-08-05 Instapaper is still the best plain article reader, so switch only when what you save has stopped being articles. If your saves are a mix of videos, threads, repos and products, use Kyurations, which renders each one in the shape of its source and reads the articles among them in the app. If you want highlights, notes and review, use Readwise Reader. If you want a large tagged archive, use Raindrop.io. Instapaper strips a page to text, which is exactly what an essay wants and exactly what a YouTube video, a GitHub repo or a product page does not. The second reason people move is that a folder of stripped articles is hard to show anyone. ### Kyurations (Free) For collections that are not only articles. Best for: People saving across a dozen platforms. - Strength: Each card takes the shape of its platform, so a repo looks like a repo and a track looks like a track - Strength: The long ones still read like Instapaper: full text, own type, no page furniture - Strength: Notes, files and voice notes are first-class items, not attachments - Strength: Publishing is one tap - Limit: No offline reading - Limit: No unread count working its way down ### Readwise Reader (Around $10 a month) Instapaper plus highlights, notes and recall. Best for: Readers who mark up what they read. - Strength: Excellent highlighting - Strength: Newsletters and PDFs in the same place - Limit: Heavier than Instapaper - Limit: Paid only ### Raindrop.io (Free, paid tier around $3 a month) For an archive rather than a queue. Best for: People with thousands of saves. - Strength: Unlimited free bookmarks - Strength: Strong search - Limit: Thin reading view ### Questions **Is Instapaper still maintained?** Yes. It has outlived most of its competitors, including Pocket, and remains the cleanest article reader available. **What is better than Instapaper for highlights?** Readwise Reader. Highlighting is the centre of the product rather than a feature inside it, and highlights export into note apps. **Can I import my Instapaper saves elsewhere?** Yes. Instapaper exports a CSV of URLs and folders, which Raindrop, Readwise Reader, Karakeep and Kyurations all accept. Kyurations turns each Instapaper folder into a draft list and keeps the selection you saved as the item's note. --- ## Bookmark manager alternatives to the browser bar Source: https://www.kyurations.in/alternatives/bookmark-manager Updated: 2026-08-05 Browser bookmarks break down at a few hundred links because a folder tree gives you no title, no preview and no way to search what a page was about. Replace them with Raindrop.io if you want a bigger, faster, searchable version of the same idea, with Kyurations if you want the saved links to become collections worth showing, with Readwise Reader if you read and highlight, and with Karakeep if you want to self-host. A browser bookmark stores a URL and whatever the page title happened to be that day. Six months later that is not enough to know what you saved or why. Folders force one home per link, which is the wrong shape for anything that belongs to two topics at once. Nothing in a browser bar can be shown to another person. ### Kyurations (Free) Bookmarks that become a collection instead of a pile. Best for: People who save with intent and want to show the result. - Strength: One click from the browser extension, and the card arrives complete, article body included - Strength: An item can belong to several lists at once, because most things do - Strength: Lists form themselves, then you arrange them - Strength: Publish and export whenever you want - Limit: Not a folder tree - Limit: Installs from the browser rather than from an app store, and there is no native app ### Raindrop.io (Free, paid tier around $3 a month) The browser bar, done properly. Best for: People who like folders and want them to scale. - Strength: Unlimited free bookmarks - Strength: Nested collections and tags - Strength: Apps everywhere - Limit: Shared collections read as folders ### Karakeep (Free, self-hosted) Self-hosted, open source, yours. Best for: Anyone who has been burned by a shutdown notice. - Strength: Full-text archive of saved pages - Strength: Automatic tagging - Limit: You maintain it ### Readwise Reader (Around $10 a month) For bookmarks that are really reading material. Best for: People whose links are research. - Strength: Highlighting and review - Strength: Everything in one queue - Limit: Paid only ### Why folders stop working A folder makes you decide, at the moment of saving, the one place a link belongs. That decision is cheap when you have forty links and impossible when you have four hundred, because by then most things belong in two places and you cannot remember which one you chose. Tags fix part of it. Automatic tags fix more of it, because the tag you would have typed is rarely the tag you type six months later. ### Questions **What is the best bookmark manager in 2026?** Raindrop.io for storage and search, Kyurations for collections you want to publish, Readwise Reader for reading and highlighting, Karakeep for self-hosting. There is no single winner, because the four are built for different jobs. **Can I import my browser bookmarks?** Chrome, Firefox, Safari, Edge and Arc all export an HTML bookmarks file, and Raindrop, Readwise Reader, Karakeep and Kyurations read it. Kyurations rebuilds your bookmark folders as draft lists, so the structure survives the move without any of it arriving as a collection you are supposed to have curated. **Are bookmark managers free?** Raindrop.io and Kyurations both have real free tiers. Karakeep is free if you host it. Readwise Reader and Pinboard are paid. --- ## Readwise Reader alternatives, including the free ones Source: https://www.kyurations.in/alternatives/readwise-reader Updated: 2026-08-03 Readwise Reader is the strongest read-and-highlight tool available, and the usual reason to look elsewhere is the price. If you wanted the reading, Instapaper does it for less. If you wanted the saving and organising, Raindrop.io does it free. If you wanted a collection you could publish rather than a private archive, Kyurations does that free and reads the articles in the app while you highlight them. Nothing else matches Reader's highlight review, so if that is why you came, pay for it. Reader bundles three products: a reader, a highlighter with review, and a save-everything inbox. Most people only use one of them heavily. At around $10 a month it is the most expensive tool in this category, which makes the question of which part you use worth answering honestly. ### Instapaper (Free, paid tier around $6 a month) If you came for the reading. Best for: People who read long articles on a phone. - Strength: Cleanest article view - Strength: Real offline reading - Limit: Highlights are basic - Limit: No review system ### Raindrop.io (Free, paid tier around $3 a month) If you came for the inbox. Best for: People who save far more than they read. - Strength: Unlimited free saves - Strength: Fast search and tags - Limit: No review, thin reading view ### Kyurations (Free) If the archive was supposed to become something. Best for: People who want to publish what they collect. - Strength: Reads articles and PDFs in the app, and highlighting happens while you read, in one of four pens - Strength: Marks stay painted in the article, and clicking one in the panel jumps back to it - Strength: Highlights and clipped regions attach to the item rather than becoming new entries, and each quote links back to the exact words on the source - Strength: Any list publishes as a page under your handle, highlights included - Limit: No spaced review of highlights - Limit: No newsletter inbox, and no direct sync into Obsidian, Notion or Roam ### Questions **Is there a free Readwise Reader alternative?** Not one that replaces all of it. Raindrop.io covers the saving free, Instapaper covers the reading on a free tier, and Kyurations covers reading, highlighting and publishing free. The highlight review system has no free equivalent, and that is the part worth paying for. **Do I need Readwise as well as Readwise Reader?** Reader includes Readwise. The older Readwise product is the highlight review layer, and a Reader subscription covers both. **Can I export my Readwise highlights?** Yes, to Markdown, CSV and directly into Obsidian, Notion and Roam. Export quality is one of the genuinely good things about it. --- ## Omnivore alternatives after the ElevenLabs shutdown Source: https://www.kyurations.in/alternatives/omnivore Updated: 2026-08-03 ElevenLabs acqui-hired the Omnivore team in October 2024 and the service closed on 15 November 2024, with roughly two weeks to export before the data was deleted. Nothing replaces it exactly, because Omnivore was open source, free, and a genuinely good reader all at once. Karakeep is the closest on philosophy and you host it yourself, Readwise Reader is the closest on highlight review and costs around $10 a month, and Kyurations is the closest on the reading itself: free, hosted, and the whole article rendered in the app from a capture taken in your own browser. Omnivore was open source, which is exactly why the shutdown stung. People had chosen it to avoid this outcome. The code is still on GitHub and a community fork runs at omnivore.work, so self-hosting is a real option rather than a theoretical one. The hosted service is gone and the data with it. If you exported in that two-week window you have an archive; if not, you do not. ### Karakeep (Free, self-hosted) Open source, self-hosted, and actively maintained. The philosophical successor. Best for: People who picked Omnivore because it was open source. - Strength: No company that can be acqui-hired out from under you - Strength: Full-text search over archived pages, with automatic tagging - Strength: Docker install and a healthy release cadence - Limit: You are the operations team, including backups - Limit: The reading experience is not as polished as Omnivore's was ### Readwise Reader (Around $10 a month) The closest match on features, if you will pay for it. Best for: People who used Omnivore's highlights and its newsletter inbox. - Strength: Highlighting, notes and spaced review - Strength: Newsletters, PDFs, YouTube and articles in one queue - Strength: Exports into Obsidian, Notion and Roam - Limit: Closed source, which is the thing you were avoiding - Limit: No free tier ### Kyurations (Free) For the part of Omnivore that was building a library, not clearing a queue. Best for: People whose saves were a collection with a shape. - Strength: Reads in the app: the whole article in its own type, PDFs included, and it picks up where you stopped - Strength: The extension captures the article from the page you are already signed in to, so a paywalled piece arrives whole - Strength: Highlight as you read, and each quote links back to the exact words on the source - Strength: Cover, author, tags and a gist arrive with the save - Strength: Any list publishes as a page, which Omnivore never did - Limit: Not open source - Limit: No offline reading, and no unread count to work down ### Wallabag (Free self-hosted, or around 11 euros a year hosted) The older self-hosted read-later, still going. Best for: People who want the Omnivore model with a longer track record. - Strength: Has outlasted most of this category - Strength: Real mobile apps and browser extensions - Strength: Imports from Pocket, Instapaper and Omnivore exports - Limit: The interface looks its age - Limit: No highlight review ### Questions **Why did Omnivore shut down?** ElevenLabs hired the founding team in October 2024 to work on its ElevenReader text-to-speech app. The product was not acquired, it was discontinued, and the service closed on 15 November 2024. **Can I still self-host Omnivore?** Yes. The codebase is still on GitHub and a community fork runs at omnivore.work. It is unmaintained by the original team, so you are adopting an orphan, but it works. **What is the best free Omnivore alternative?** Karakeep or Wallabag if you can host it, Kyurations if you want something hosted and free without a self-hosting setup. Readwise Reader is the closest on features but is paid only. **Is there a free Omnivore alternative that reads articles in the app?** Kyurations does, free and hosted. The whole article renders in the app and you highlight as you read. There is no offline mode and no unread count, so it is a library rather than a queue you work down. Karakeep and Wallabag also read in-app if you are willing to host them. **Can I recover my Omnivore data?** Only from an export taken before 15 November 2024. Everything else was deleted. --- ## Evernote Web Clipper alternatives for saving links Source: https://www.kyurations.in/alternatives/evernote-web-clipper Updated: 2026-08-05 Most people leaving the Evernote Web Clipper are leaving the price rather than the clipper, which is still one of the better ones. If you want the full-page capture and the archive, Raindrop.io does it free and Karakeep does it self-hosted. If you want the saved pages to become collections you can publish, and to read the clipped article in the app, use Kyurations. If your clips were really research notes, Notion or Obsidian with a clipper is a better shape than another bookmark tool. Evernote's pricing and device limits pushed a lot of long-term users out, and the clipper is usually the last thing they replace because it worked. The clipper did two different jobs: archiving whole pages, and dropping references into a notes system. Those need different replacements. ### Raindrop.io (Free, paid tier around $3 a month) The closest free replacement for the clipping and the archive. Best for: People who mainly want their pages kept and findable. - Strength: Unlimited free bookmarks with previews - Strength: Permanent page copies on the paid tier, which is the Evernote archive behaviour - Strength: Imports an Evernote export - Limit: Not a notes app - Limit: Shared collections look like folders ### Kyurations (Free) For clips that were always meant to become something you show. Best for: People whose clippings were a subject collection, not a filing system. - Strength: Cover, author, platform, tags and a gist filled in automatically - Strength: The article body is kept and reads in the app, so a clipped piece survives the page going down - Strength: Highlights and clipped regions attach to the page instead of becoming new notes - Strength: Publish any list as a page - Limit: The article text is kept, not a full-page snapshot with its layout - Limit: Not a replacement for Evernote's notebooks ### Obsidian with a web clipper (Free for personal use) If the clips were the raw material for writing. Best for: People who were using Evernote as a second brain. - Strength: Local Markdown files nobody can take away - Strength: Official web clipper since 2024 - Strength: Links between notes, which Evernote never did well - Limit: You build the system yourself - Limit: Sync costs extra ### Karakeep (Free, self-hosted) Self-hosted archiving, no subscription. Best for: People who want the archive and want to own it. - Strength: Full-text search over archived pages - Strength: Automatic tagging - Limit: You run the server ### Questions **Does the Evernote Web Clipper still work?** Yes. The clipper is not the problem people are solving for; the pricing and device limits are. **How do I move my Evernote clips somewhere else?** Export as ENEX from Evernote. Raindrop, Obsidian and Notion all import it. Kyurations does not read ENEX, so the route there is a CSV or an HTML list of the source URLs, which it does read; each one is re-fetched with a current cover and tags as it arrives. **What is the best free alternative to Evernote's clipper?** Raindrop.io for archiving and search, Kyurations for collections you want to publish, Obsidian if the clips feed your own writing. All three have real free tiers. --- ## Pinboard alternatives when you want more than a URL Source: https://www.kyurations.in/alternatives/pinboard Updated: 2026-08-03 Pinboard is fast, cheap and intentionally featureless, so people leave it for one of two reasons: they want to see what they saved, or they want more than one person maintaining it. Raindrop.io answers both and is free. Kyurations answers the first and adds publishing and an in-app reader. Karakeep answers the second by making you the maintainer. If neither reason is yours, stay: nothing here is faster. Pinboard is a one-person operation, which is both why it is still cheap and still fast, and why some people want a fallback. The other reason is visual. A list of blue links is hard to scan once it runs into the thousands, and impossible to show anyone. ### Raindrop.io (Free, paid tier around $3 a month) Pinboard with covers, and a team behind it. Best for: People who want the same archive, legible. - Strength: Unlimited free bookmarks - Strength: Previews, nested collections and a fast search - Strength: Imports a Pinboard export directly - Limit: Heavier than Pinboard - Limit: Page archiving is on the paid tier ### Kyurations (Free) For when the tag pile was supposed to be a collection. Best for: People who want to show someone the good ones. - Strength: Cards that look like their source instead of a row of URLs - Strength: Tags reconciled automatically rather than typed by hand every time - Strength: Saved articles read in the app, and the text survives the original page going down - Strength: Publish a list as a page under your handle - Limit: The article text is kept, not a full-page snapshot - Limit: Newer, with less history behind it ### Karakeep (Free, self-hosted) Own the server, own the fallback. Best for: People whose worry is bus factor. - Strength: Open source - Strength: Full-text archive and automatic tagging - Limit: You are now the one-person operation ### linkding (Free, self-hosted) The closest thing to self-hosted Pinboard. Best for: People who want Pinboard's exact shape, on their own box. - Strength: Tiny and fast, same spirit as Pinboard - Strength: Imports Pinboard and Netscape HTML - Limit: Minimal by design, so no previews - Limit: Small community ### Questions **Is Pinboard still maintained?** Yes, by its founder, as it has been since 2009. That longevity is the point of it, and the single-maintainer risk is the reason some users keep a mirror elsewhere. **Can I export my Pinboard bookmarks?** Yes, as JSON or Netscape HTML. Raindrop, linkding, Karakeep and Kyurations all take one or both. **What is the best free Pinboard alternative?** Raindrop.io if you want it hosted, linkding or Karakeep if you want to host it yourself. Pinboard itself is paid, at around $22 a year. --- ## mymind alternatives that cost less than $12 a month Source: https://www.kyurations.in/alternatives/mymind Updated: 2026-08-03 mymind's pitch is that you save anything and never organise it, and it costs around $12 a month. If the visual, everything-in-one-place feel is what you want without the price, Kyurations is free and does automatic tags and gists at save time. If you want the organising back, Raindrop.io is free and folder-shaped. Nothing free matches mymind's interface polish, and that is worth being honest about. mymind is deliberately opinionated: no folders, no tags you type, no sharing. That is the appeal and, for some people, the wall they hit. The two reasons people look elsewhere are the price and the fact that nothing in it can be shown to anyone. ### Kyurations (Free) The same save-anything-and-it-sorts-itself idea, free, and publishable. Best for: People who liked mymind but wanted to show someone the result. - Strength: Automatic tags, cover, author and a short gist on every save - Strength: Lists form themselves rather than being filed by hand - Strength: Images, notes, files and links all first-class - Strength: Publish a list as a page, which mymind does not do - Limit: Less visually polished than mymind, which is a genuinely high bar - Limit: Installs from the browser rather than from an app store, and there is no native app ### Raindrop.io (Free, paid tier around $3 a month) For when you actually did want folders. Best for: People whose mymind grew past the point of finding anything. - Strength: Unlimited free saves - Strength: Nested collections, tags, strong search - Limit: Nothing automatic about the organising - Limit: Plainer to look at ### Are.na (Free tier, around $7 a month for unlimited) For visual collecting that is meant to be seen by others. Best for: Designers and researchers building reference boards. - Strength: Channels are public and connected by design - Strength: A real community around the collections - Limit: Free tier caps monthly blocks - Limit: No automatic metadata or tagging ### Questions **Is there a free alternative to mymind?** Kyurations is free and closest on the save-anything-and-it-sorts-itself idea. Raindrop.io is free if you want folders back. Neither matches mymind's interface polish. **Can I share things from mymind?** No, and that is deliberate. mymind is built as a private space. If sharing is what you need, that is the clearest reason to use something else. **How much does mymind cost?** Around $12 a month, or roughly $100 a year, as of July 2026. Check their site, because pricing moves. --- ## Are.na alternatives for visual collecting Source: https://www.kyurations.in/alternatives/are-na Updated: 2026-07-27 Are.na is the best tool for building public reference channels, and people usually look elsewhere for one of two reasons: the free tier's monthly block limit, or the fact that saving something takes several deliberate steps. Kyurations is free with no block limit and fills in the metadata for you. Raindrop.io is free if you want a private archive rather than a public channel. Nothing else has Are.na's community, which is often the actual reason to stay. Are.na's free tier limits how many blocks you can add per month, which is the usual first wall. The second is friction: Are.na is built for deliberate connection-making, which is the point, and is slower than a one-click save. ### Kyurations (Free) One-click saving, automatic metadata, and lists you can publish. Best for: People who want the public collection without the monthly cap. - Strength: No limit on how much you save - Strength: Cover, author, tags and a gist arrive without typing - Strength: Lists form themselves and publish as pages under your handle - Limit: No community browsing what you build, which is Are.na's whole point - Limit: Not a board, so no spatial arrangement ### Raindrop.io (Free, paid tier around $3 a month) If the channels were really a private archive. Best for: People collecting reference for themselves. - Strength: Unlimited free saves - Strength: Fast search across everything - Limit: Nothing public or social about it ### Milanote (Free tier, around $10 a month unlimited) If what you wanted was the board, spatially. Best for: Designers arranging moodboards. - Strength: Free-form canvas - Strength: Good for visual thinking in progress - Limit: Free tier caps notes and images - Limit: Not built for collecting from the web ### Questions **Is Are.na free?** There is a free tier with a monthly limit on new blocks, and an unlimited plan at around $7 a month. The limit is the most common reason people look at alternatives. **What is the closest thing to Are.na?** Nothing matches its community, which is the real product. For the collecting itself, Kyurations is free and faster to save into, and Milanote is closer if you want a spatial board. **Can I export my Are.na channels?** Yes, Are.na has a public API and export options, and channel contents come out as JSON. --- ## Chrome bookmarks alternatives once the bar runs out Source: https://www.kyurations.in/alternatives/chrome-bookmarks Updated: 2026-08-05 Chrome's bookmark manager stores a URL, a title and a folder, and that is all it has ever done. It stops working at a few hundred links because a title from a page you skimmed once tells you nothing later. Move to Raindrop.io for a bigger searchable archive, to Kyurations if the links are a collection you would show someone, to Readwise Reader if they are reading material, or to Karakeep if you want to host it yourself. Raindrop, Readwise Reader, Karakeep and Kyurations all read a Chrome bookmarks export; Kyurations rebuilds your folders as draft lists, so the structure survives without anything arriving as a collection you are supposed to have curated. Chrome bookmarks sync between your own devices and nowhere else, so nothing you save can be shown to another person. One link, one folder. That decision is fine for forty links and impossible for four hundred. No previews, no notes, no automatic tags. Six months on, a bookmark is a title you do not recognise. ### Raindrop.io (Free, paid tier around $3 a month) The bookmark bar, scaled up properly. Best for: People who like folders and want them to keep working. - Strength: Unlimited free bookmarks with covers - Strength: Nested collections, tags and fast search - Strength: Apps on every platform, so it is not tied to one browser - Limit: Shared collections read as folders - Limit: Organising is still manual ### Kyurations (Free) For links that were meant to become a collection. Best for: People who save with intent and want the result to look like it. - Strength: Cover, author, platform, tags and a gist arrive with the save - Strength: One item can live in several lists, because most things do - Strength: Anything long enough to read opens in the app rather than sending you back out - Strength: Publish any list as a page - Limit: Not a folder tree - Limit: Installs from the browser rather than from an app store, and there is no native app ### Readwise Reader (Around $10 a month) If most of your bookmarks are things to read. Best for: People with a reading backlog rather than a link archive. - Strength: Best-in-class highlighting and review - Strength: Articles, PDFs and newsletters in one queue - Limit: Paid only - Limit: Private by design ### Karakeep (Free, self-hosted) Self-hosted, so no vendor to outlive. Best for: People who run their own things. - Strength: Open source - Strength: Full-text search over archived pages - Limit: You maintain and back it up ### Exporting your Chrome bookmarks Open chrome://bookmarks, use the menu at the top right, and choose Export bookmarks. You get a single HTML file containing every link, folder and date. That file is the universal format for this category. Every tool listed here reads it, and so does Safari, Firefox and Edge if you ever go back. ### Questions **What is the best replacement for Chrome bookmarks?** Raindrop.io if you want the same idea done properly and free, Kyurations if the links are a collection you would publish, Readwise Reader if they are a reading list, Karakeep if you want to self-host. **How do I export bookmarks from Chrome?** Go to chrome://bookmarks, open the menu at the top right, and choose Export bookmarks. It saves one HTML file, which Raindrop, Readwise Reader, Karakeep and Kyurations all read. In Kyurations you drop it into Settings and your folders come back as draft lists. **Will I lose my bookmarks if I switch?** No. Export first, import into the new tool, and leave the Chrome copy where it is until you are sure. Nothing about importing removes the originals. **Is there a bookmark manager that works across browsers?** All four here do, because they store your links on their own service rather than inside one browser profile. That is usually the second reason people move. --- ## Kyurations vs Raindrop.io Source: https://www.kyurations.in/vs/raindrop-io Updated: 2026-08-05 Pick Raindrop.io if you want a large, searchable archive of links in nested folders, and pick Kyurations if you want the things you save to become collections other people can look at. Raindrop is the better filing cabinet and has years more polish behind it. Kyurations is the better display case: cards take the shape of their source, lists form themselves out of what you save, articles read in the app rather than sending you back out, and publishing a list is one tap. Two differences are harder to close than the rest: your collection is addressable by Claude, Cursor and other agents over MCP, and an article saved through the extension is captured from the browser you are already signed in to, so a paywalled piece arrives whole rather than as a stub. Your Raindrop export moves across as one file. Both are free to start. ### Kyurations compared with Raindrop.io - Price: Kyurations Free; Raindrop.io Free, paid tier around $3 a month - Organising model: Kyurations Semantic lists that form themselves, many per item; Raindrop.io Nested folders and tags, one home per link - What arrives on save: Kyurations Cover, author, platform, tags, short gist; Raindrop.io Cover and title, tags by hand - How a saved item looks: Kyurations A card shaped like its platform; Raindrop.io A uniform row or grid tile - Publishing: Kyurations Any list becomes a page under your handle; Raindrop.io A shared collection, folder-shaped - Notes and files: Kyurations First-class items alongside links; Raindrop.io Attachments on the paid tier - Reading: Kyurations Articles and PDFs open in the app, with your place kept; Raindrop.io Sends you to the source - How the article is fetched: Kyurations Captured in your own browser, so a paywalled piece arrives whole; Raindrop.io Fetched server-side - Import: Kyurations Reads a Raindrop CSV or HTML export, collections become draft lists; Raindrop.io Reads most bookmark files - Ask across your collection: Kyurations Ask Kyuri, with citations to your items; Raindrop.io No - Agent access: Kyurations An MCP endpoint, so Claude and other agents can read the collection; Raindrop.io No - Page archiving: Kyurations The article text is kept, not a full-page snapshot; Raindrop.io Full page copies, on the paid tier - Mobile apps: Kyurations Installs from the browser, plus the extension; Raindrop.io iOS, Android, macOS, Windows - Export: Kyurations Any time, free; Raindrop.io Any time, free Choose Kyurations when: - Your saves are meant to be shown to someone eventually - You would rather not name and file every link by hand - You save across platforms, so a video, a repo and a product should not all render as the same grey row - You want to ask questions across your whole collection and get answers with citations back to your own items Choose Raindrop.io when: - You think in nested folders and want them many levels deep - You have tens of thousands of links and search speed is the whole product - You need iOS and Android apps today - You want a permanent snapshot of the whole page, layout and all, which Raindrop's paid tier does and Kyurations does not ### The real difference in one sentence Raindrop assumes you will come back and find the link. Kyurations assumes someone else will come and look at the collection. Every other difference follows from that. Folders serve retrieval. Cards, covers and self-forming lists serve display. ### Questions **Is Kyurations free like Raindrop?** Yes, and there is no gated tier holding back the features described here. Export is free too. **Can I move from Raindrop to Kyurations?** Yes, in one file. Raindrop exports CSV and HTML and Settings reads either, with your collections arriving as draft lists rather than finished ones, so nothing lands in a collection pretending you curated it. Links land in seconds and their covers, authors and tags fill in afterwards. What is left to weigh is not the move, it is whether you want three thousand undifferentiated bookmarks inside a tool built for the ones you would show someone. **Which has better search?** Raindrop, on raw speed across a very large archive. Kyurations answers questions across your collection instead, which is a different thing and better when you cannot remember the words you saved. --- ## Kyurations vs Pocket Source: https://www.kyurations.in/vs/pocket Updated: 2026-08-05 Pocket no longer exists. Mozilla closed it on 8 July 2025 and deleted user data after 8 October 2025, so this is a comparison with a memory rather than a product. Pocket did two jobs: a read-later queue and a pile of things you liked. Kyurations takes the second job seriously and does the reading without the queue, so an article opens in the app but no unread count is kept and there is no offline mode. If the number going down is what you miss, take Instapaper or Readwise Reader instead. ### Kyurations compared with Pocket - Status: Kyurations Live; Pocket Shut down 8 July 2025 - Price: Kyurations Free; Pocket Was free with a paid tier - Reading view: Kyurations Yes, articles and PDFs in the app; Pocket Yes - Ask across what you saved: Kyurations Ask Kyuri, with citations to your items; Pocket No - Read-later queue: Kyurations No, deliberately; Pocket Yes, the core of it - Offline reading: Kyurations No; Pocket Yes - Automatic tags and gist: Kyurations Yes; Pocket Tags by hand - Non-article saves: Kyurations Videos, repos, products, notes, files; Pocket Mostly articles - Publishing a collection: Kyurations Yes, one tap; Pocket No - Import: Kyurations Reads the Pocket CSV or HTML export you kept; Pocket n/a - Agent access: Kyurations An MCP endpoint, so Claude and Cursor can read and curate it; Pocket No - Export: Kyurations Any time; Pocket Window closed 8 October 2025 Choose Kyurations when: - The saves were never really a queue, they were a collection - You want covers, authors and tags without typing them - You want to publish the good parts rather than let them sit unread Choose Pocket when: - You wanted the unread count and the sense of a backlog draining, which Kyurations deliberately has no version of - You read on a phone, on a plane, without a connection ### Questions **Can Kyurations import my Pocket export?** Yes. Pocket exported both a CSV and an HTML file and Settings reads either one, keeping the titles and tags it carries. Every link is saved straight away and its cover and summary arrive afterwards. Pulling out only what you still care about is still the better instinct for a read-later archive, but that is now your call rather than our limitation. **Does Kyurations have a read-later queue?** No, and that is a decision rather than a gap. Articles and PDFs do read in the app, and it keeps your place in them, but nothing counts what you have not got to yet. A second pile that nags you would compete with the collection for attention. If the queue is what you want, Instapaper and Readwise Reader are the honest recommendations. --- ## Kyurations vs Notion for saving links Source: https://www.kyurations.in/vs/notion Updated: 2026-08-05 Notion is a better workspace and Kyurations is a better place to keep links. A Notion database can store bookmarks, but the clipper saves a title and a URL and leaves you to fill in every other property by hand, which nobody keeps up for long. Kyurations fetches the cover, author, platform, tags and a short gist at save time. Stay in Notion when the link belongs inside a project with other data around it. ### Kyurations compared with Notion - Price: Kyurations Free; Notion Free tier, paid plans per seat - What the clipper saves: Kyurations Cover, author, platform, tags, gist; Notion Title and URL - Structure: Kyurations Lists that form themselves; Notion Databases you design - Formulas and relations: Kyurations No; Notion Yes - How saves look: Kyurations Platform-shaped cards; Notion Table rows or gallery cards you configure - Publishing: Kyurations A page under your handle; Notion A public database view - Collaboration: Kyurations Share and subscribe; Notion Full multiplayer editing - How the article is fetched: Kyurations Captured in your own browser, so a paywalled piece arrives whole; Notion Title and URL only - Import: Kyurations Reads a Notion CSV export of links; Notion Reads CSV into a database - Agent access: Kyurations An MCP endpoint over your collection; Notion An MCP server over your workspace Choose Kyurations when: - The links are the thing, not a property of something else - You are not going to fill in database columns by hand, and you know it - You want the collection to be publishable as a page rather than a shared database view Choose Notion when: - The links live inside a project, next to tasks, docs and dates - You need formulas, relations or rollups over your saves - Your team already works in Notion and a second tool is friction ### Questions **Why is the Notion Web Clipper so limited?** It builds a database row rather than parsing a link. Notion has no opinion about what a saved page should contain, so it captures the title and the URL and leaves the rest of the properties empty. **Can I use both?** Most people should. Project material stays in Notion, the collection lives where links are first-class, and both export cleanly. --- ## Kyurations vs Readwise Reader Source: https://www.kyurations.in/vs/readwise-reader Updated: 2026-08-05 Readwise Reader is the better tool if reviewing what you highlighted is the whole job, and it costs around $10 a month. Kyurations is free, reads articles and PDFs in the app as well, and is better at turning what you save into a collection someone else can look at. It also captures the article in your own browser rather than fetching it server-side, so a paywalled or login-walled piece arrives whole, and it exposes your collection to Claude and Cursor over MCP. If your saves are research that ends in your own notes, pay for Reader. If your saves are finds that end in a list you send to someone, use Kyurations. ### Kyurations compared with Readwise Reader - Price: Kyurations Free; Readwise Reader Around $10 a month - Reading view: Kyurations Yes, articles and PDFs, with your place kept; Readwise Reader Yes, strong - How the article is fetched: Kyurations Captured in your own browser, so a paywalled piece arrives whole; Readwise Reader Fetched server-side, so some paywalled pages arrive partial - Highlights: Kyurations Made while reading in four pens, painted in the article, shown on public lists; Readwise Reader Central, with spaced review - Spaced review: Kyurations No; Readwise Reader Yes, the core of it - Ask across what you saved: Kyurations Ask Kyuri, with citations to your items; Readwise Reader Search and filters, no chat over your library - Offline reading: Kyurations No; Readwise Reader Yes - Non-text saves: Kyurations First-class; Readwise Reader Supported, text-shaped - Automatic tags and gist: Kyurations Yes; Readwise Reader Partial - Publishing: Kyurations Any list, one tap; Readwise Reader No - Note-app export: Kyurations Full export, no direct sync; Readwise Reader Direct sync to Obsidian, Notion, Roam Choose Kyurations when: - You want it free - Your saves are not mostly text: videos, repos, tracks, products - The end state is a published collection rather than a note file Choose Readwise Reader when: - You highlight heavily and want spaced review of old highlights - You read newsletters and PDFs in the same queue as articles - You read offline, on a plane or a commute - Your highlights need to land in Obsidian, Notion or Roam ### Questions **Does Kyurations do highlights?** Yes, and you make them while reading the article in the app, in one of four pens. A highlight or a clipped region attaches to the page you already saved rather than creating a new entry, it stays painted in the article when you come back, clicking it in the panel jumps to it, it links back to the exact words on the source, and it shows on a published list because the quotes you pulled are part of the curation. What there is no version of is Readwise's spaced review. **Is there spaced review of highlights?** No. That is Readwise's strength and there is no equivalent here. **Does Kyurations have a reading view?** Yes. Articles and PDFs open in the app, in its own type, and it remembers where you stopped. Saved through the browser extension, the article is captured from the page you are already signed in to, so a paywalled or login-walled piece arrives whole rather than as a stub. There is no offline mode. --- ## Kyurations vs Instapaper Source: https://www.kyurations.in/vs/instapaper Updated: 2026-08-05 Instapaper is the better tool if what you save is articles you intend to read offline or on a phone, and its reading view is still the cleanest in this category. Kyurations reads articles and PDFs in the app as well, and is the better tool if what you save is a mix of videos, repos, tracks, products and notes that you want arranged and shown. Instapaper strips a page to text, which is right for an essay and wrong for almost everything else. ### Kyurations compared with Instapaper - Price: Kyurations Free; Instapaper Free, paid tier around $6 a month - Reading view: Kyurations Yes, articles and PDFs, with your place kept; Instapaper Yes, the best there is - How the article is fetched: Kyurations Captured in your own browser, so a paywalled piece arrives whole; Instapaper Fetched server-side - Offline reading: Kyurations No; Instapaper Yes - Unread count: Kyurations No, deliberately; Instapaper Yes - Non-article saves: Kyurations Videos, repos, tracks, products, notes, files; Instapaper Articles, mostly - How a save looks: Kyurations A card shaped like its platform; Instapaper Stripped text - Automatic tags and gist: Kyurations Yes; Instapaper No - Ask across what you saved: Kyurations Ask Kyuri, with citations to your items; Instapaper No - Publishing: Kyurations Any list, one tap; Instapaper A public folder link - Highlights: Kyurations Made while reading, shown on public lists; Instapaper Yes, on the paid tier - Import: Kyurations Reads the Instapaper CSV, folders become draft lists; Instapaper n/a - Agent access: Kyurations An MCP endpoint, so Claude and Cursor can read and curate it; Instapaper No Choose Kyurations when: - Your saves are not mostly articles - You want the cover, author and tags filled in rather than stripped out - The collection is meant to be published, not worked through Choose Instapaper when: - You read long articles, and you read them on a phone - You need offline reading on a plane or a commute - You want an app that does nothing but reading - You want an unread count that goes down, which Kyurations does not have ### Questions **Can I use both?** Plenty of people do, and it is a sensible split. Articles you intend to read go to Instapaper, everything you want to keep and arrange goes to Kyurations. **Does Kyurations have a reading mode?** Yes. An article or a PDF opens in the app, set in its own type with the page furniture gone, and it remembers where you stopped. Saved through the browser extension, the text is captured from the page you are already signed in to, so a paywalled piece arrives whole. What Instapaper still does better is offline reading and the unread count, and neither of those is coming. --- ## Kyurations vs Are.na Source: https://www.kyurations.in/vs/are-na Updated: 2026-08-05 Are.na is the better choice if you want your collecting to sit inside a community that browses and connects to it, and the deliberate friction of adding a block is part of what makes those channels good. Kyurations is the better choice if you want to save in one click, have the cover, author and tags filled in for you, and publish without a monthly cap. The real question is whether you want an audience that is already there, or a page you send to someone. ### Kyurations compared with Are.na - Price: Kyurations Free, no cap; Are.na Free with a monthly block limit, around $7 a month unlimited - Saving: Kyurations One click, metadata fetched; Are.na Deliberate, metadata by hand - Community: Kyurations None yet; Are.na Its strongest feature - Arrangement: Kyurations Lists, ordered; Are.na Channels and blocks, spatial - Automatic tags and gist: Kyurations Yes; Are.na No - Publishing: Kyurations A page under your handle; Are.na A public channel - API and export: Kyurations Full export; Are.na Public API and export - Agent access: Kyurations An MCP endpoint, so Claude and Cursor can read and curate it; Are.na A public API, no MCP endpoint Choose Kyurations when: - You want no monthly limit on what you save - You would rather not type the title, the tags and the source every time - You save from many platforms and want each to look like itself - You want to ask Kyuri questions across everything you have collected, and get answers that cite your own items Choose Are.na when: - The community is the point, and you want people to find your channels - You think spatially and want blocks you arrange on a board - You want other people's channels connected to yours ### Questions **Is Kyurations trying to be Are.na?** No. Are.na is a community with a collecting tool inside it. Kyurations is a collecting tool where publishing is one tap. If the community is what you want, Are.na is the answer and there is no substitute for it. **Which is better for moodboards?** Are.na, or Milanote if you want a free-form canvas. Kyurations arranges lists, not boards. --- ## Raindrop.io vs Pocket vs Kyurations Source: https://www.kyurations.in/vs/raindrop-vs-pocket Updated: 2026-08-05 Pocket closed on 8 July 2025, so this is really a two-way comparison with a ghost. Raindrop.io took most of Pocket's users and is the strongest free bookmark archive available. Kyurations is the one to pick if the pile was always meant to become a collection you publish rather than an archive you search, and it reads a saved article in the app rather than sending you back out. If what you miss about Pocket is the unread count and offline reading, neither of these is it, and Instapaper or Readwise Reader is. ### Kyurations compared with Raindrop.io and Pocket - Status: Kyurations Live; Raindrop.io and Pocket Raindrop live, Pocket closed July 2025 - Price: Kyurations Free; Raindrop.io and Pocket Raindrop free, paid tier around $3 a month - Organising: Kyurations Lists that form themselves, many per item; Raindrop.io and Pocket Folders and tags, filed by hand - What arrives on save: Kyurations Cover, author, platform, tags, gist; Raindrop.io and Pocket Cover and title - Reading: Kyurations Articles and PDFs open in the app; Raindrop.io and Pocket Pocket read well, Raindrop sends you out - Read-later queue: Kyurations No, deliberately; Raindrop.io and Pocket Pocket had one, Raindrop does not really - Publishing: Kyurations A page under your handle; Raindrop.io and Pocket A shared folder - Import: Kyurations Reads either export, folders become draft lists; Raindrop.io and Pocket Raindrop reads most bookmark files - How the article is fetched: Kyurations Captured in your own browser, so a paywalled piece arrives whole; Raindrop.io and Pocket Fetched server-side - Agent access: Kyurations An MCP endpoint, so Claude and Cursor can read and curate it; Raindrop.io and Pocket No - Mobile apps: Kyurations Installs from the browser, plus the extension; Raindrop.io and Pocket Raindrop: everywhere Choose Kyurations when: - The pile was supposed to become something other people see - You would rather not name and file every link by hand - You save across platforms and want a video to look like a video Choose Raindrop.io and Pocket when: - You want the biggest free archive with the fastest search - You think in nested folders - You need iOS and Android apps today - You want permanent copies of saved pages, on Raindrop's paid tier ### Questions **Is Raindrop.io the best Pocket replacement?** For the saving and tagging, yes. For the reading, no. Pocket was doing both jobs, and Raindrop only does the first one well. **Which is better, Raindrop or Kyurations?** Raindrop if your links are an archive you will search. Kyurations if they are a collection you will show. That is the whole difference, and everything else follows from it. --- ## Kyurations vs browser bookmarks Source: https://www.kyurations.in/vs/browser-bookmarks Updated: 2026-08-05 Browser bookmarks are free, instant and already installed, and for under about a hundred links there is no reason to use anything else. They break at scale for three specific reasons: one link can only live in one folder, nothing is stored except a URL and whatever the page title happened to be, and nothing you save can be shown to another person. Kyurations fixes all three and costs nothing, at the price of being one more place to save. ### Kyurations compared with browser bookmarks - Price: Kyurations Free; browser bookmarks Free - What is stored: Kyurations Cover, author, platform, tags, gist, notes; browser bookmarks URL and page title - Homes per link: Kyurations As many lists as fit; browser bookmarks One folder - Search: Kyurations Across titles, tags, notes and gists; browser bookmarks Titles and URLs - Sharing: Kyurations Publish a page under your handle; browser bookmarks Not possible - Across browsers: Kyurations Yes; browser bookmarks Only within one browser's sync - Import: Kyurations Reads the HTML bookmarks file, folders become draft lists; browser bookmarks n/a - Ask across what you saved: Kyurations Yes, answers cite your own items; browser bookmarks No - Agent access: Kyurations An MCP endpoint, so Claude and Cursor can read and curate it; browser bookmarks No Choose Kyurations when: - You cannot remember what a bookmark from March was about - Things keep belonging in two folders - You want to send someone the good ones - You save videos, threads and repos as well as pages Choose browser bookmarks when: - You have under a hundred links and can still find them - You want zero new accounts and zero new habits - Everything you save is a tool you use daily, not something you will revisit once ### Questions **Do I have to give up my browser bookmarks?** No. Export the HTML bookmarks file and Settings reads it, rebuilding your folders as draft lists you sign or discard. Nothing about importing removes the originals, so most people bring the pile across, keep the daily tools in the bookmark bar, and let the ones worth showing rise out of it. **How many bookmarks is too many for a browser?** The number at which you stop opening the manager, which for most people is a few hundred. The count matters less than whether you can still find anything. --- ## What content curation actually is Source: https://www.kyurations.in/writing/what-is-content-curation Updated: 2026-07-27 Content curation is choosing a small number of things from a large number, arranging them so the order and the grouping carry meaning, and putting the result where someone can see it. All three parts are required. Saving links is collecting, and collecting on its own produces an archive rather than a collection. The difference is that a curator leaves things out. ### The three parts Selection is the part people skip. A curator's real work is rejection: deciding that most of what crosses their desk does not belong. A folder with four thousand links in it has had no selection applied to it at all, which is why nobody, including its owner, ever opens it. Arrangement is the part people underrate. The same twelve links in a different order are a different argument. Put the strange one first and the collection reads as provocation. Put it last and it reads as a reward. Museums have known this for two centuries and bookmark managers have never once acted on it. Display is the part that makes the other two matter. A collection nobody sees is a private preference, and a private preference is not curation. This is not about audience size. Showing one person is display. ### Why the tools got it wrong Almost every tool built for saving links optimises for retrieval. Search, folders, tags, full-text archives. All of it answers the question of how to find the thing again. That is a real question, and it is not the interesting one. Nobody is short of ways to find a link. What people lack is a reason for the link to have been kept, and something to do with it once it has. The result is a category of software where the most common outcome is a large pile that its owner feels vaguely guilty about. ### The bowerbird test A male bowerbird builds a structure out of collected objects and arranges them by colour and size. It steps back, looks at its own work, and moves things. The bower has no function beyond being looked at. That is the test for whether a collection is curated. Did anyone step back and look at it? Did anything move as a result? Was it built to be seen? Most saved-link piles fail all three. ### What good curation looks like in practice - Small. A list of nine things you stand behind beats a list of ninety you half-remember. - Ordered. Not by date added, which is an accident, but by something you chose. - Annotated. One line on why this one is here does more work than any tag. - Named. A list called Reading is not a collection. A list called Things that changed how I write is. - Finished enough to send. Not perfect. Sendable. ### Questions **What is the difference between content curation and content aggregation?** Aggregation gathers everything that matches a rule, automatically and at scale. Curation selects a few things by judgement and arranges them. An algorithmic feed is aggregation. A reading list with nine entries and a reason for each is curation. **Is saving bookmarks the same as curating?** No. Saving is collecting, which is the first third. Without selection, arrangement and display you have an archive, and an archive is a tool for finding things again rather than a thing anyone looks at. **How many items should a curated collection have?** Fewer than you want to include. There is no correct number, but the useful discipline is that adding a twelfth item should mean removing one, because a collection that only grows has stopped being curated. **Does curation require an audience?** It requires display, which is not the same as an audience. A collection you would send to one person passes. A collection you would not show anyone is a private preference. --- ## Why saved links die Source: https://www.kyurations.in/writing/why-saved-links-die Updated: 2026-07-27 Saved links die because saving is free and returning is expensive. At the moment you save, you pay nothing and feel like you have done something. Months later the link has no cover, no context and no reason attached, so opening it costs more than ignoring it. The fix is to make the save carry its context, and to give the pile somewhere to go. ### The save is a promise you make to a stranger The person who will open that link is not you. It is a version of you six months out who has forgotten the article you had just read, the argument you were having, and the reason this seemed worth keeping. A bare URL and a page title tell that stranger almost nothing. So the stranger does what anyone does with an unlabelled box in a garage. They leave it there. ### Three failures, in order of how much damage they do - No context at save time. Title and URL only, so the why is gone the moment you close the tab. - One home per link. Folders force a decision you cannot make correctly, because most things belong in two places. - No destination. The pile has no state it can reach where it counts as done, so it only grows, and a thing that only grows becomes a thing you avoid. ### What actually fixes it Make the save complete itself. If the cover, the author, the platform, the tags and a one-line gist arrive on their own, the stranger six months out has enough to decide in a second rather than in a click. Let one thing belong to several places, because it does. The decision you avoid at save time is the decision that stops you saving. Give the pile a destination. A published list is a finish line. Once a collection can be sent to someone, the pile stops being a backlog and starts being material. ### The uncomfortable part Most of what you saved was never worth keeping. That is fine, and it is the normal ratio for anyone who reads a lot. Curation is mostly deletion. The pile does not need to be processed. It needs to be cut down to the nine things that were actually good, and those nine need somewhere to be seen. ### Questions **Why do I never read my saved articles?** Because saving is free and reading is not. The save gives you the feeling of having acted, which discharges the urge that would otherwise make you read it now. This is well documented in how people use read-later tools, and it is why read-later queues grow rather than drain. **Should I delete my old bookmarks?** Most of them, yes. A useful pass: keep anything you can state a reason for in one line, and drop the rest. It will be a small fraction, and the small fraction is the collection. **How do I make saved links useful again?** Give them context at save time, allow one item in several lists, and pick a destination for the good ones. A list you would send to a specific person is the most reliable destination, because it forces selection. --- ## How to organise bookmarks so you use them Source: https://www.kyurations.in/writing/how-to-organize-bookmarks Updated: 2026-08-04 Organise bookmarks in four passes, then keep maintenance near zero. Cut anything you cannot give a one-line reason for, which will be most of them. Group what survives by subject rather than by intent, because Read later is not a subject. Let anything that belongs in two groups sit in both. Then publish the two or three groups that are actually good, because a collection with a destination stays organised and a collection without one does not. ### Steps 1. Cut first. Delete anything you saved for a reason you can no longer remember. Most of a bookmark pile is dead weight, and sorting dead weight is the slowest way to do nothing. 2. Group by subject, never by intent. Fold everything into subjects you would say out loud. Folders named after intentions, To Read or Someday, fill up and are never opened. 3. Allow two homes. Stop forcing one item into one folder. A thing that genuinely belongs in two places should sit in both, or you will hesitate every time you save. 4. Give the good ones somewhere to go. Pull the handful worth keeping into a list you would show someone. A collection with an audience gets maintained; an archive does not. 5. Keep maintenance near zero. Do nothing on a schedule. Tidy the one subject you are already working in, when you are already in it. ### Pass one: cut Open the pile and keep only what you can justify in a single line. Not a summary, a reason. This one changed how I think about pricing. This one is the best explanation of the algorithm I have found. Expect to keep somewhere between five and fifteen per cent. That is not failure, that is the normal hit rate for anything you saved impulsively. ### Pass two: group by subject, never by intent Read later, To do and Important are intents, and intents expire. Six months on, everything is Read later, so the group means nothing. Subjects survive. Typography, Pricing, Cold email, Sea creatures. You can tell from the name whether a thing belongs, which is the entire job of a group name. - If a group has one item, it is not a group yet. Leave the item ungrouped. - If a group has more than about thirty items, it is two groups. - If you cannot name a group in under four words, you have not found the subject. ### Pass three: allow two homes An essay about pricing psychology belongs in Pricing and in Behavioural economics. Forcing it into one is how you lose it, because you will look in the other one. Any tool that makes you pick a single folder is making you do work that will be wrong half the time. Tags, or membership in several lists, fix this. ### Pass four: give the good ones somewhere to go Pick the two or three groups you would actually show someone and publish them. Send one to a specific person. This is the step that keeps the system alive. A group that other people can see acquires standards, and standards are what stop a collection from silting up again. Nothing else you do will maintain itself. ### Maintenance, which should be almost nothing - Save with context, so nothing arrives bare and needing work later. - Review a published list when you add to it, not on a schedule. - Delete on sight. A link you scroll past twice without opening is a link you do not want. ### Questions **What is the best way to organise bookmarks?** By subject, with items allowed in more than one group, and with the best groups published so they have a reason to stay good. Organising by intent, such as Read later or Important, fails within months because the intent stops being true. **Folders or tags for bookmarks?** Tags, or any model that lets one item live in several places. Folders force a single choice at save time that is wrong for anything belonging to two subjects, which is most things worth keeping. **How often should I clean up my bookmarks?** Almost never, if saves arrive with their context attached and you delete on sight. Scheduled cleanups fail because they are a chore with no reward. Publishing works because it has one. **How many bookmarks is too many?** The number where you stop opening the app. For most people that is a few hundred. The count matters less than whether any group in it is finished enough to show someone. --- ## How to share a curated list people actually open Source: https://www.kyurations.in/writing/how-to-share-a-curated-list Updated: 2026-07-27 A shared list gets opened when it is short, named for what it is about, ordered on purpose, and annotated one line at a time. Nine items is a good ceiling. A link dump with thirty entries and a name like Resources gets closed in four seconds, because nothing in it tells the reader where to start or why any of it was worth their attention. ### Steps 1. Name it for its subject. Title the list after what is in it, not the container it lives in. Resources tells a reader nothing they can act on. 2. Cut to nine. Take out everything you would not defend. A list of nine chosen things is read; a list of forty is skimmed and closed. 3. Order it on purpose. Put the entry that proves the point first. Order is the cheapest editing you can do and the most visible. 4. Write one line each. Say what the thing is and why it is here. One honest line per entry is the whole difference between a collection and a pile of links. 5. Send it to one person. Share it with somebody specific rather than posting it at everyone. A list written for one reader is better than a list written for an audience. ### Name it for its subject, not its container Resources, Links and Useful stuff describe the box. Nobody opens a box. Name the argument instead. The nine essays that made me rewrite my landing page. Everything I read before hiring my first engineer. Now the reader knows within a second whether this is for them. ### Cut to nine Thirty items reads as a search result. Nine reads as a decision. The reader can feel the difference, and the feeling is what makes them start. If you cannot get under about a dozen, you have two lists. ### Order it on purpose Date added is not an order, it is an accident of your attention. Choose something: easiest to hardest, most surprising first, the one that changed your mind at the end. Say the order out loud in the description. In rough order of how much each one annoyed me. That single line does more for the reader than any tag. ### One line each Not a summary. A reason. Read the first three paragraphs and stop. The tables are the point. He is wrong about the conclusion and right about everything before it. This is the part that cannot be automated, and it is the part that makes the list yours. ### Send it to one person Not to an audience. To a person, by name, with a sentence about why you thought of them. Lists made for a specific reader are better than lists made for everyone, and they are the ones that get forwarded. ### Questions **How many links should a shared list have?** Under a dozen, and nine is a good target. Longer lists read as search results rather than as recommendations, and the reader stops choosing. **What makes someone open a shared list?** A name that states the subject, a length they can see the end of, and a first line that says why the list exists. In that order. **Should I annotate every link?** Yes, with one line, and a reason rather than a summary. The summary is already on the page you are linking to. The reason is the only thing you are adding. --- ## Bookmarks and read-later are not the same thing Source: https://www.kyurations.in/writing/bookmarks-vs-read-later Updated: 2026-07-27 A bookmark is a permanent reference you expect to return to, like a tool or a document. A read-later save is a temporary intention that should disappear once acted on. They have opposite lifecycles: bookmarks accumulate on purpose, read-later items are supposed to drain. Putting both in one place is the most common reason people end up with thousands of saved links and no working system. ### Opposite lifecycles A bookmark is supposed to still be there in three years. The tax calculator, the API reference, the invoice template. Growth is fine, because you are building a reference shelf and a shelf is meant to fill up. A read-later save is supposed to be gone in a week. You saw an essay, you did not have twenty minutes, you saved it. The save has done its job when you read it or admit you never will. One pile that only grows and one pile that should drain, sharing a folder, is why the whole thing feels bad. The permanent items make the temporary ones look like debt, and the temporary ones bury the permanent ones. ### The third pile nobody names There is a category that is neither. Things you keep because they were good: an essay that changed your mind, a project you want to remember, a piece of work you admire. You are not going to return to it for reference and you are not going to work through it as a queue. You keep it because it says something about what you value, and the only sensible thing to do with it is arrange it and show someone. That third pile is what curation is for. It is the one that most tools have no answer to, which is why it ends up dumped into the other two. ### What to do about it - Reference goes in the browser bar or a bookmark manager. Small, flat, and boring on purpose. - Read-later goes in a reader with an unread count, so it can drain. Instapaper or Readwise Reader. - The good stuff goes into collections you would show someone. That pile earns arrangement. - When a read-later item turns out to be excellent, promote it into the third pile. That is the only migration path that matters. ### Why one tool cannot do all three It is not a feature gap, it is a conflict about what a save means. A read-later app is right to nag you about unread items. A reference tool would be insufferable if it did. A collection wants to be seen; a reference shelf does not care. Tools that try to serve all three end up with an inbox, a folder tree and a share button that nobody uses, which is roughly the history of this category. ### Questions **What is the difference between a bookmark and a read-later app?** A bookmark is a permanent address you expect to reuse. A read-later save is a temporary intention that should disappear once you read it or give up on it. Bookmarks are meant to accumulate; read-later items are meant to drain. **Should I use one app or two?** At least two, and arguably three. Reference in a bookmark manager, reading in a reader with an unread count, and things you keep because they were good in a collection you would show someone. **Why does my read-later list keep growing?** Because saving is free and reading is not, and the save discharges the urge that would otherwise make you read it now. A queue with no cost to adding will always grow. The fix is deleting on sight rather than reading faster. --- ## How to build a reading list people actually follow Source: https://www.kyurations.in/writing/how-to-build-a-reading-list Updated: 2026-07-27 A reading list gets followed when it has a narrow subject, a stated point of view, a visible standard for what gets in, and a rhythm someone can rely on. The most common failure is breadth: a list called Interesting Links has no subject, so nobody can predict what turns up, and unpredictable is the same as unfollowable. Pick one subject you would still be reading about in two years. ### Steps 1. Narrow the subject. Pick a question narrow enough to finish. A list about everything you are curious about never reaches a second entry. 2. State the standard out loud. Write down what earns a place before you add anything. An unstated standard drifts until the list means nothing. 3. Give every entry a reason. One line per item, and make it the reason it is there rather than a description of what it is. 4. Choose rhythm over volume. Add on a cadence you can hold. A list that gains one good thing a week outlives one that gained thirty in an afternoon. 5. Show your working when you change your mind. When you remove something, say why. A list that visibly revises itself is worth more than one that only ever grows. ### Narrow beats broad, every time Interesting Links is not a subject. Neither is Tech, Design or Business. If a reader cannot predict roughly what the next entry will be about, there is nothing to follow. The test: could someone describe your list to a friend in one sentence, without using the word interesting? Pricing pages that actually explain their pricing. Essays by people who changed their minds in public. Everything I read while learning to sail. Narrow feels like a limit and works like a promise. ### State the standard out loud Write down what gets in and what does not, and put it at the top of the list. Nothing that is mostly a product announcement. Nothing I could not finish. Nothing behind a hard paywall. This does two things. It tells a reader what your judgement is worth, and it makes your own decisions faster, because most candidates fail a stated standard immediately. ### One line per entry, and make it a reason Not a summary. The summary is on the page you are linking to and adds nothing. A reason is the only thing you are contributing. Skip to section three, the rest is throat-clearing. This is the only honest thing I have read about churn. He is wrong about the conclusion and right about the method. This is the part that cannot be automated and the part people follow you for. ### Rhythm over volume Three good entries a month beats thirty in one week and nothing after. A reader is subscribing to a habit, and an irregular habit is one they will stop checking. Pick a cadence you can hold in a bad month, not a good one. ### Show your working when you change your mind Removing something you previously recommended, and saying why, is the strongest signal of judgement a list can send. Almost nobody does it. A list that only ever grows is an archive. A list where things get taken out is a curated one, and readers can feel the difference even when they cannot name it. ### Questions **How often should I update a public reading list?** On a rhythm you can hold in a bad month. Three entries a month, reliably, beats thirty in a burst followed by silence. Readers subscribe to a habit more than to a volume. **How many links should a reading list have?** For a one-off list you send someone, under a dozen. For a running list people follow, there is no ceiling, as long as entries are added at a predictable rate and old ones get removed. **What makes a reading list worth following?** A narrow subject, a stated standard for what gets in, one line of reasoning per entry, and a predictable rhythm. Breadth is the most common reason a list has no followers. **Should I explain why each link is included?** Yes, in one line, and give a reason rather than a summary. The summary is already on the linked page. The reason is the only thing you are adding to it. --- ## What Kyurations costs, and what it does not do Source: https://www.kyurations.in/writing/is-kyurations-free Updated: 2026-08-03 Kyurations is free. There is no paid tier gating the features described anywhere on this site, no limit on how much you save, and you can export your entire collection at any time. Curators can put a price on a list they publish, so a collection somebody else built may cost money, but building and publishing your own never does. It also deliberately does not do several things: there is no unread count, no offline reading, no full-page snapshot archive, and no native mobile app. If any of those is what you came for, another tool is the better answer. ### What free means here - No cap on saved items, lists or published collections. - Automatic cover, author, tags and gist on every save, not held back for a paid tier. - The reader is included. Articles and PDFs open in the app, it keeps your place, and you highlight as you read. - Publishing, subscriptions on your lists, and the browser extension, all included. - Full export whenever you want it. Your collection leaves in the same shape it arrived. ### Where money does change hands A curator can put a price on a list they publish and sell it. If you buy one, you are paying that person for their collection, and Kyurations takes a cut of the sale. That is the only place a payment appears. Nothing you do with your own library is behind it. Saving, reading, highlighting, arranging and publishing are free whether or not you ever price a list. ### What it does not do There is no read-later queue. Articles do read in the app and it remembers where you stopped, but nothing counts what you have not got to and no number goes down. That is a decision: a pile that nags you competes with the collection for attention. Instapaper and Readwise Reader are the right tools if you want the count. There is no offline reading. The article is stored, but you need a connection to open it. There is no full-page snapshot. The article text is kept, so a piece survives its page going down, but the layout around it is not. Raindrop's paid tier and Karakeep both keep the whole page. There is no native mobile app. It installs from the browser and works on a phone, but it is not in the app stores. ### How it stays free The honest answer is that it is early and the costs are small. There is no advertising, nothing is sold to data brokers, and your collection is not training material. If a paid tier arrives, it will be for things that cost real money to run rather than for features that already exist. Anything you can do today, you will still be able to do. ### Questions **Is Kyurations free?** Yes. There is no paid tier, no cap on how much you save, and no feature described on this site is held back behind a subscription. **If it is free, why can some lists be bought?** Because the price is set by the curator, not by us. Anyone publishing a list can choose to sell it, and buying one pays that person for their collection. Using Kyurations to save, read, arrange and publish your own is free either way. **Can I read articles inside Kyurations?** Yes. An article or a PDF opens in the app, set in its own type, and it remembers where you stopped. Saved through the browser extension, the text is captured from the page you are already signed in to, so a paywalled or login-walled piece arrives whole. There is no offline mode. **Can I export my collection?** Yes, at any time and in full. Export is not gated and never has been. **Does Kyurations sell my data or train on it?** No. There is no advertising, nothing is sold to data brokers, and private collections are not training material. **Is there a mobile app?** Not in the app stores. It installs to a phone home screen from the browser and runs as its own app from there, and the browser extension covers desktop saving. **Will it always be free?** Everything you can do today will stay available. If a paid tier arrives it will cover things that cost real money to run, not existing features. --- ## How to chat with your bookmarks Source: https://www.kyurations.in/writing/how-to-chat-with-your-bookmarks Updated: 2026-08-03 To chat with your bookmarks you need a tool that reads what you saved rather than the web, so retrieval runs over your own items and the answer cites them. In Kyurations this is Ask Kyuri, and it is free. Save enough that there is a corpus to search, ask by meaning rather than by keyword, check the cited cards before you trust a claim, narrow to a single list when an answer comes back too broad, and turn a good answer into a list so the next person does not have to ask again. ### Steps 1. Save enough to have something to ask. A library of twenty links has nothing to tell you. The useful threshold is roughly the point where you can no longer remember everything in it, which for most people is a few hundred items. 2. Ask by meaning, not by keyword. Describe the argument rather than guessing the title. Retrieval runs over what a piece is about, so 'the video that argued taste comes from proximity, not training' finds the thing whose title you never knew. 3. Read the citations before you trust the answer. Every claim should open the card it came from. If a sentence has no card behind it, treat it as the model talking rather than your library, and check it. 4. Narrow to one list when the answer is too broad. Scope the question to a single list and the same query gets sharper, because the corpus it searches got smaller and more coherent. 5. Turn the answer into something you keep. Ask it to make a list out of what it just found, or to tag those items. Otherwise you will run the same query again in three months. ### Why a general chatbot cannot do this Ask ChatGPT what you saved about a subject and it will tell you about the subject. It has never seen your library. What comes back is the average of what has been written, which is exactly the thing you were trying to escape when you started saving the good ones. The value in a collection is the judgement in it. Every save is a small decision that this was worth keeping, and a few hundred of those decisions is a filter nobody else has. An assistant answering out of that corpus is answering out of your taste. An assistant answering out of the open web is answering out of everyone's. ### What good questions look like - Recall: the piece you remember the shape of but not the name of. - Synthesis: what four saved articles actually disagree about. - Stock-taking: how much you have saved on a subject, and whether it is enough to write from. - Resurfacing: what you saved a year ago on something you are thinking about today. ### What it will not do It cannot answer from things you never saved. That sounds obvious and is the single most common disappointment: a library is a small corpus, and a question outside it should come back as 'nothing here covers that' rather than as a confident paragraph. It is not a search engine with better manners either. If you know the exact title, search is faster and always will be. Chat earns its place on the questions where you cannot name what you are looking for. ### Questions **Can I chat with my bookmarks?** Yes, if the tool holding them reads them. Kyurations does this through Ask Kyuri: it searches your saved items by meaning, answers from them, and cites the ones it used so you can open each source. It is free and needs no set-up. **How is this different from asking ChatGPT the same question?** ChatGPT answers from what it was trained on. Ask Kyuri answers from what you decided was worth keeping, and shows which of your items each part of the answer came from. On a question about your own collection, the small curated corpus wins. **How many bookmarks do I need before this is useful?** Enough that you cannot hold them all in your head, which is usually a few hundred. Below that you can still scroll the list faster than you can ask a question about it. **Can it organise my library as well as answer questions?** Yes. It can create a list from what an answer found, add or remove items, tag them, save a link or a note, and publish a list. Deletes are a recoverable soft delete, and every write is one you asked for in the thread. --- ## The case for an AI that reads your library, not the web Source: https://www.kyurations.in/writing/ai-that-reads-your-library Updated: 2026-08-03 An assistant grounded in your own saved library answers better than a general one on any question about what you have read, because the corpus was filtered by a human with a stake in it. A few hundred deliberate saves encode judgement that no amount of training data reproduces, and grounding an answer in them makes it checkable: every claim opens the item it came from. The trade is coverage. A curated corpus cannot answer what you never saved, and a tool that pretends otherwise has given up the only advantage it had. ### A small corpus is the feature The instinct is that more data is better, and for training a model it is. For answering a question about what you have read it is the opposite. Every item in your library got there because you decided it was worth keeping, and that decision is a signal the open web does not carry. So the corpus is small, opinionated and yours, and those are the three properties that make the answers feel unlike a chatbot's. It is not smarter. It is better informed about one specific thing, which is what you have been paying attention to. ### Grounding is what makes it checkable An answer with no sources is a claim you have to take on trust, and the failure mode of language models is that a wrong answer looks exactly like a right one. Citations change the shape of the problem: if every sentence opens the card it came from, verifying is a click rather than a research project. This also constrains what the tool can get away with. When references are index based, pointing at a real retrieved item, a fabricated citation cannot resolve to anything. The honest answer when the library does not cover a question is that it does not. ### Where it fails, and should say so Coverage is the trade. Ask about something outside your library and there is nothing to retrieve, and the right behaviour is to say so rather than to fall back on general knowledge and present it in the same voice as a cited answer. A tool that blurs those two has thrown away the reason to use it. The other limit is that a library reflects what you saved, including its gaps and its enthusiasms. If you saved eleven articles making the same argument, a summary will report a consensus that exists in your collection and not in the world. Being answered out of your own taste is the point, and it is also the bias. ### What this looks like in practice - Retrieval first, then generation, so the answer is assembled from real items rather than recalled. - Citations that resolve to cards you can open, not footnote-shaped decoration. - Full text where it exists, so a question about a long report is answered from the report. - The ability to act on a result, because an insight you cannot file is one you will look for again. ### Questions **Why not just use a general AI assistant?** Use one, for most things. The narrow case where a library-grounded assistant wins is any question about what you personally have read and saved, because a general model has never seen your collection and cannot cite it. **Does answering from my library mean my collection trains a model?** No. Private collections are not training material. Retrieval means the items are fetched to answer your question in the moment, not absorbed into a model's weights. **What happens when my library does not cover the question?** You should be told that. An assistant that quietly switches to general knowledge while still sounding cited is worse than one that has no answer, because you can no longer tell which mode you are in. --- ## What to ask a library you built Source: https://www.kyurations.in/writing/questions-to-ask-your-library Updated: 2026-08-03 The questions worth asking a collection you built are the ones you cannot answer by scrolling: what you half remember but cannot name, what several saved pieces disagree about, how much you have gathered on a subject, and what you saved a year ago on something you are thinking about now. The questions that waste your time are the ones a search box answers faster, and anything the collection does not contain. A library is a small corpus, so its strength is depth on what you kept, not coverage of what you did not. ### The four that work - Recall: 'the essay that argued saving is a substitute for reading.' You remember the argument and not the title, which is precisely where a keyword search fails and a meaning search does not. - Synthesis: 'what do the pieces I saved on pricing actually disagree about?' Useful because it forces the collection to speak as a set rather than as a list. - Stock-taking: 'how much have I saved on this, and when?' The answer tells you whether you have enough to write from or are still collecting. - Resurfacing: 'what did I save about hiring last year?' Old saves are invisible by default, and this is the only reliable way back to them. ### The ones that waste your time Anything where you know the exact title. Search is faster, and asking a question to get one link back is a slower search box. Anything your library does not contain. The honest answer is that nothing covers it, which is correct and still not what you wanted, so it is worth knowing before you ask. Anything that needs the whole library at once, such as 'summarise everything.' A collection of a few thousand items has no single summary, and a tool that produces one is generalising rather than retrieving. ### Ask, then put it away The step people skip is the last one. A good answer is a piece of work: it found six items, worked out how they relate, and put them in an order. If you close the thread, that work is gone and you will do it again. Turning the answer into a list, or tagging what it found, is what makes the second question cheaper than the first. Over time the questions you ask most become the lists you keep, which is the collection organising itself around how you actually use it. ### Questions **What should I ask my saved links first?** Something you half remember and cannot name. It is the question a folder tree and a search box both fail, so it is the fastest way to find out whether asking your library is worth doing at all. **Why does asking beat searching?** It does not, in general. Search wins whenever you know the words that are in the thing. Asking wins when you know what a piece argued but not what it was called, or when the answer lives across several items rather than in one. **Can I ask about one list rather than everything?** Yes, and you should when a broad answer comes back vague. A smaller, more coherent corpus produces a sharper answer, which is the same reason a well-kept list is more useful than a large pile. --- ## Keeping things on purpose Source: https://www.kyurations.in/writing/keeping-things-on-purpose Updated: 2026-07-28T05:43:02.747+00:00 People have gathered objects, images and ideas that mattered to them across every recorded culture. The tools have changed; the instinct has not. A collection is always an argument, and the arrangement says something the individual object cannot say alone. ### Questions **Who made the earliest known curated collection?** Ennigaldi-Nanna, a Babylonian princess and priestess at Ur, assembled objects from earlier civilisations around 530 BCE and labelled them with small clay tags explaining their origins. By any reasonable definition, she was running a museum. **What is a commonplace book?** A commonplace book was a personal collection of copied passages from letters, sermons, poems and other texts, organised by theme and annotated over years. John Locke developed a formal indexing system for his, but many ordinary people kept them too as a kind of second memory built by hand. **What is the difference between saving links and actually collecting?** A folder of ten thousand unread links is a pile, not a collection. Collecting requires a decision about what to keep, some arrangement that lets pieces speak to each other, and some return to and use of the material. **Why do collections always reflect their maker?** Every collection is shaped by decisions about what to include and what to exclude, and those decisions carry values whether or not the person making them has named those values aloud. The boundary drawn around a collection is always partial, always personal, and always a portrait of whoever made it. ---