# Changelog All notable changes to this project will be documented in this file. The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/), and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html). ## [0.9.1074] ### Changed - **The task-list header gets out of the way while you read.** On a small phone the Tasks header — title, search field, saved-filter rail and active filter chips — took up to half the screen before the first task appeared. Scrolling down now folds it into a slim bar with the title and two buttons, so the reclaimed space shows two to three more tasks; a deliberate upward scroll (or a tap on the bar's title, which carries a chevron to say it's tappable) brings the full header back — an accidental thumb jiggle doesn't. The same fold applies wherever the list is narrow: a narrowed desktop window, and the list pane of the desktop split view. The slim bar still says *what* is narrowing the list, not just that something is: an ad-hoc filter is named as a count of its clauses ("3 filters"), an active saved view shows its name next to the title, and an active search shows the query in quotes. The search button reopens the header with the field focused, the filter button opens the filters directly, and applying changes there reopens the header so the chips confirm what you did. Typed search text survives the fold, a list too short to scroll never collapses, and with reduced motion enabled the switch is instant instead of animated. - **An unfiltered task list no longer wears chips.** The chip row above the list used to echo the default status set (Open, Groomed, In Progress) as removable chips even when nothing was narrowed, making the resting view look filtered. Status chips now appear only when the selection actually deviates from the default open-work view, and once two or more things are narrowing the list — clauses, the search query, or both — the row ends with a "Clear all" chip that restores the default open-work view and clears the query, so leaving a filter session is one tap instead of one per chip. - **Numbers in the Tasks header now mean one thing.** The saved-filters button in the mobile rail carried the number of *saved* filters, which sat right next to counts that mean "how many tasks match" and read as "1 filter is active" on an unfiltered list. The button now carries no number; the saved-filters sheet itself shows what exists. - **The saved-views button is called Views.** It opens saved views, but it was labelled "Filters" — sitting directly beside the funnel that opens the actual filters, so the two most-used controls in the header shared a name and led to different places. - **Colour in the filter chips means what the chip means.** Every active filter chip sat on the same tinted plate regardless of what it stood for, so a red P0 priority chip was drawn on a green background. The plate is now neutral and the chip's own colour carries the meaning. - **An ad-hoc filter keeps its name on a narrow screen.** On a 375-point phone the "Custom" pill was squeezed until only its task count was left, so the one control naming your current filter showed a bare number. The save affordance now yields its label first. ### Fixed - **The "no tasks match" message no longer sticks to the screen edge.** When the Tasks list's empty-state message wrapped to a second line — a longer language such as German, or a larger system text size — it was laid out across the full width of the screen, flush against the left edge with no margin. It now uses the same centered, inset empty-state treatment as the Logbook, with the tasks glyph above it. ## [0.9.1073] ### Added - **The manual keeps every released version readable.** The manual's version dropdown now lists each published release next to Development, every release keeps its exact pages and screenshots permanently, and the manual's front door leads to the newest release once one exists. Older manuals also learn about releases published after them, so no snapshot ever points at a stale list. ### Fixed - **Planning after your working day has ended now finishes instead of retrying an impossible answer.** Daily OS was told to add no blocks once the planning window closed, but its required final step rejected an empty plan. It can now save an explicit empty draft, keep any existing blocks intact, and surface selected work that no longer fits instead of looping until the request fails. - **Daily OS plans stay inside your configured working hours.** On an overloaded day, the planner could keep the total duration under your capacity by leaving a gap and then scheduling the final task after your working day ended. Plans outside the configured window are now rejected and retried, so work must be shortened or omitted explicitly instead of quietly overrunning. - **Assistant proposals are in your language.** Every suggested change — set a title, change a status, add a checklist item, log a time entry, link two tasks — was written in English at the moment the assistant thought of it, and then shown to you in English no matter what language the rest of the app was in. Those rows are now composed as you read them, so they appear in your own language, with the relationship between two tasks phrased properly rather than pasted together word by word. Suggestions your device does not recognise — from a newer version, say — still show their original wording rather than nothing at all. - **The German and Spanish relationship pickers say the direction they mean.** Offering to link two tasks, the German dropdown read "Duplikat von" — "duplicate of" — for the direction that actually records the opposite: the current task being duplicated *by* the chosen one. Spanish had the same problem with follow-ups, where "Tiene seguimiento de" reads as tracking the other task rather than gaining it as a follow-up. They now read "Wird dupliziert von" and "Tiene como seguimiento", matching what picking them does. - **The sync diagnostics panel is translated.** Its heading, its refresh button and its "Last ignored" list stayed English on every device, whatever language the rest of the app was set to. - **The sync statistics counters say the right thing in five more languages.** "DB Missing Base" counts entries held back waiting on a record that has not arrived yet, but the Spanish, French and Romanian translations said *the database is missing* — contradicting the legend on the very same panel — and the Italian and Dutch ones stacked the words in an order that read as neither. The Romanian buttons on that page also addressed you informally, unlike the rest of the app. - **The refresh button on Backfill sync is easier to hit, and says what it is.** It was a 16-pixel icon with barely any room around it — a 24-point target, half the size a finger needs — so on a phone it took an accurate tap. The icon is unchanged, but the button now occupies a full touch target, which makes the header of that card taller. Screen readers also named it without ever saying it was a button. - **The Italian theming settings were labelled "Il nome" ("The name").** The Italian catalog mistranslated the Theming settings entry and its logging domain; both now read "Tema", and the mode labels use the standard "Aspetto scuro" / "Aspetto chiaro" wording. The Dutch and Swedish light-mode labels ("Lichtschijn", "Ljusets utseende") were similarly garbled and now read "Licht uiterlijk" and "Ljust utseende". - **The privacy policy described device pairing incorrectly.** It stated that encryption keys are exchanged via the QR code and never transmitted over the network. Neither half was true: the QR code carries your sync account credentials, while the encryption keys travel between devices over the homeserver — encrypted, and only to devices you have emoji-verified. The practical consequence had gone unsaid too: anyone who photographs the pairing code can sign in to your sync account and lock you out, even though they still cannot read your entries. The policy now says all of this, in the same words the manual already used. - **The privacy policy promised a data export that does not exist.** It offered to export your data "at any time". There is no such feature, and none is needed: your logbook is an ordinary SQLite file on your own disk that any SQLite client can read. The policy now explains how to take a consistent copy and where the attachments live, rather than pointing at a button that was never there. - **Reporting a security vulnerability no longer starts by publishing it.** The security policy asked for vulnerability reports in the public issue tracker, which the contributing guide separately tells you not to do, and which exposes everyone running Lotti before a fix exists. Reports now go through GitHub's private vulnerability reporting. ### Changed - **The Matrix Stats page reads as one page again.** Its headings had been set by hand and disagreed with each other — "Sent messages" shouted while "Top KPIs" was left at whatever size it inherited — so the page looked like three unrelated panels stacked together. Every group heading now sits at the same level, the counter tiles share one rhythm, and the numbers on them line up with the figures on Backfill sync. - **The app description says what Lotti actually tracks.** It now mentions that AI usage is recorded per category and model — tokens and requests for every cloud call, and cost, energy and CO₂e wherever the provider reports them — and that planning and time recording stay separate so intent and outcome remain distinct facts. It no longer describes third-party providers as "GDPR-compliant", because that is their claim to make rather than Lotti's, and no longer lists time recording twice. - **The documentation no longer promises impact figures every provider can supply.** The README and privacy policy described cost, energy and CO₂e as recorded for every AI call. Only providers that return those numbers with the response are measured — today Melious — while everyone else contributes token counts and leaves the rest empty rather than estimated. Local inference is not measured at all. The same pages also described a just-in-time notice naming the provider before each AI feature runs, which does not exist: routing is chosen in settings, and once a category is configured its work can reach that provider without a further prompt. - **The documentation stops calling a freshly paired device a backup.** Pairing gives a new device everything written from then on, but existing settings and earlier entries arrive only when *Send settings* and *Send message history* are run from a device that has them. Describing every paired device as a complete replica could have led someone to retire their only full copy. The docs now say what pairing does and does not carry, and no longer describe the journal as an append-only log — editing an entry overwrites it rather than adding a version. ## [0.9.1072] ### Added - **Models can be removed right where a provider lists them.** A provider's detail page shows its models and offers "Add model", but taking one away meant switching to the Models tab and finding the same model again in the full list. Each model row on the provider page now carries a small trash action; a tap asks for confirmation and then removes the model, with the same brief undo window available for deletions elsewhere in the app. ### Fixed - **A confirm button no longer disappears when its label is long.** The buttons at the bottom of a dialog were placed side by side or stacked according to how wide the screen was, without ever measuring the buttons themselves. A longer translation — German and Romanian most often — therefore still counted as fitting on a phone: the confirm button was squeezed down to nothing and a striped overflow warning ran across the footer, leaving the dialog with no visible way to confirm it. Dialog footers now measure the actual buttons and give each its own line as soon as they stop fitting side by side. - **Lotti installs on older iPhones and iPads again.** The app had come to require iOS 17, which left devices that cannot be updated past iOS 15 — the iPhone 6s and the first-generation iPhone SE among them — unable to install it at all. Nothing in the app actually needed iOS 17: the requirement arrived with a speech feature that was removed from iOS shortly afterwards, and was never lowered again. Lotti now asks for iOS 15.8, the last iOS 15 release, so those devices are supported once more as long as they are fully up to date. ### Changed - **Core screens now fit the smallest supported phones.** Agent wake rows, AI usage figures, Daily OS planning and refinement, speech transcripts, session ratings, and What's New no longer overflow or cut off their controls at the original iPhone SE's 320 × 568 layout size. Short forms and planning panels scroll when necessary, while compact rows wrap or shorten safely. - **Sync lists line up with the rest of the app on narrow phones.** The Sync Conflicts and Outbox pages gave their lists a one-off side margin on screens narrower than most phones; they now use the same spacing step as every other list, so the two pages no longer sit two pixels tighter than their neighbours. ## [0.9.1071] ### Added - **Encryption keys are never shared with unverified devices — and sync never stops for them.** Keys now go only to devices you have emoji-verified; an unverified device receives ciphertext it cannot read. Until now the app protected you the blunt way: one unverified device — typically a dead session left by an uninstalled app — silently stopped all syncing on every device until it was removed. Now everyone you trust keeps syncing, the device list warns that the unverified device can't read new entries, and verifying or removing it stays a calm cleanup step instead of an emergency. - **See and manage every device in your sync account.** The sync status page now lists all sessions — not just unverified ones — with a "This device" marker, Verified/Unverified chips, when the server last saw each device, and a hint when a device has been silent long enough to be presumed dead. Sessions signed in with this account can be removed — except the one you're on — including previously verified ones and sessions that never finished pairing. Devices paired the old way, each with its own separate account, can be verified but not removed. While any device is excluded from sync, the list says so in plain words. - **Tell the task assistant how tasks relate — in either direction.** Saying "this task is blocked by X", "this supersedes the old migration task", or "this duplicates Y" now becomes a reviewable suggestion that records the actual relationship — blocks, follows up, duplicates, fixes, supersedes, or a plain link — between this task and an existing one. Both phrasings of a relationship work ("A blocks B" and "B is blocked by A" store the same edge), the blocked side shows up as blocked exactly as if the link were made by hand, and nothing is written until the suggestion is confirmed. - **A spoken follow-up task can carry its relationship.** "Create a follow-up task X that blocks this one" creates X and the blocking link in one confirmable suggestion, instead of a plain link that loses the "blocks" part. The assistant also sees how already-linked tasks relate to the current one, so it stops guessing parent/child roles from link direction and won't re-suggest a relationship that already exists. - **Add a device from whichever device you have to hand.** "Add device" is now an action on the device list itself, so the phone in your pocket can bring in the new laptop just as well as the other way round. The sheet shows the pairing code, tells you when the new device has actually joined, and keeps "Send settings" — for your categories, habits, dashboards and AI settings — within reach the whole time instead of below the code. Transfers unlock only after that exact new device completes emoji verification, when it can receive the encryption keys needed to read them. The sheet also says plainly what the code is: letting your own new device scan it is fine, keeping a screenshot or sending it through chat is not. - **Send message history while adding a device.** "Send message history" now sits beside "Send settings" on the existing device, where it can queue the local history the new device does not have yet. It defaults to everything, also offers the last 30 days or a custom range, and stays open with per-phase progress until the messages are queued. - **A machine with no camera can be brought in too.** The typing screen now explains how to get the code across from the other device — copy it there, and move it the way you would a password — and says so instead of doing nothing when there is nothing on the clipboard to paste. ### Changed - **Older sync maintenance screens now match the rest of the app.** Rebuilding the search index and repairing the sequence log now use the shared progress treatment, while sync statistics and this-device profile settings use the same cards, fields, rows, spacing, colors, and type as newer screens. - **Sync filters now match the rest of the app.** The Waiting, Failed, Sent, Unresolved, and Resolved controls on Sync Outbox and Sync Conflicts now use the shared design-system chip and count treatments, with consistent selected, hover, pressed, light, and dark states. - **The sync screens sit on the app's own spacing and type scales.** A handful of sizes on these pages had been picked by hand rather than taken from the app's scales, so they drifted from everything around them: the device glyph on each card in the device list was slightly undersized, the figures on Backfill sync were a shade lighter than comparable numbers elsewhere, the dotted leader lines between a label and its figure were faint enough to be hard to follow across the row, and the counter tiles in sync statistics were a little taller than wide for the same padding. All four now match the rest of the app. - **Daily OS planning now gives up on stalled or overlong AI responses sooner.** Cut-off answers no longer save partial plans; planning retries cleanly instead. - **Daily OS planning stays fast as retained history grows.** Current-day scheduling and agent wakes now use bounded database lookups instead of walking older processing jobs or captures. - **Repairing agent sync data has its own place, and runs when it matters.** The "Choose what to sync" list mixed two unrelated things: settings to send, and a repair for agent items saved without the marker other devices use to order them. The repair moved to Sync Settings → Backfill sync, and now also runs automatically just before "Send message history" sends agent data — so it no longer depends on having sent settings first. - **The sync screens follow the app's own button and layout rules again.** On the Devices page the "Add device" button is no longer an oversized bright pill, "Stop syncing this device" lines up with the cards above it instead of sitting indented, the "This device" marker no longer outshouts everything around it, and the two dates on each device card are finally the same size. In "Choose what to sync", the list was centred, which made shorter labels look like they were nested under longer ones; it now uses the same full-row selection pattern as category pickers, with labels on the left and checkboxes on the right. Confirm buttons across the sync and maintenance dialogs no longer SHOUT or stretch the full width. - **The "Enable Sync Actor" setting is gone.** The advanced-settings toggle never did anything — the isolate-based sync it was meant to switch on was never finished or connected — so it is removed rather than left as a switch with nothing behind it. - **The sync statistics legend moved onto the panel it explains.** The "Legend" button in Matrix Stats did nothing when tapped — the definitions of each counter only ever appeared as a hover tooltip, so on a phone they were unreachable. The button is gone and the definitions now sit on the "Sync Metrics" heading directly above the counters, where a tap shows them. - **A device removal the server refuses can now be finished, not just reported.** If the sync account's password was changed on another device, this one keeps syncing on its existing session but can no longer prove who it is when removing a device — and the attempt ended in a message telling you to go and do it somewhere else. It now asks for your current sync password and completes the removal on the spot, and tries to save that password so the next removal does not have to ask again. - **Verifying a device that looks dead warns you first.** The emoji check needs the other device awake, online and showing Lotti. Starting one against a device the server has not heard from in weeks left you watching a screen that would never answer. Lotti now says so before opening it, and you can still go ahead. - **The device list tells status and identity apart.** A device that has no encryption keys yet wears a quiet grey "Unverified" chip instead of borrowing the exact dress of the "This device" marker, so a glance down the list no longer reads a waiting state as an identity. And on the one card where removal is the highlighted action, the Verify button beside it now carries the accent color — it used to look like a Cancel button, when it is in fact the good way out. - **Sync cards sit on the app-wide card surface.** Every card in the sync screens now uses the same material as the rest of the app's detail pages — same tone, same hairline border, no one-off shadow — so Devices no longer looks subtly foreign next to the pages around it. - **Daily OS is opt-in again while its planning quality is improved.** Existing installs that had enabled it keep that choice, while new installs start without the Daily OS navigation entry. It can be enabled explicitly from Settings → Advanced → Config flags. - **Setting up sync starts with a real first screen.** Before anything is configured, the Devices page used to show a single row that repeated the menu entry you had just tapped, floating in an otherwise empty pane. It now says what sync is — your journal on every device you own, end-to-end encrypted, between your own devices only — with one Set up sync button. - **The pairing QR code always renders whole.** On a short window the pinned "Send settings" bar could slice the code in half — unscannable, and with nothing suggesting the rest was below. The QR and its check code now live in one pairing card that sizes itself to the window, so the two things the other device asks you to look at are always visible together — on phones too, where the pinned bar used to clip the check code. - **Every pairing screen has one clear next action.** After a failed connection the highlighted button is now Enter a new code — a rejected code usually predates a password change on the account, which only a fresh code fixes, so retrying the identical code is the quiet option beside it, not the promoted one. While the emoji ceremony is still outstanding, the paired screen shows it as a numbered gate — confirm the emoji first, receive your settings after — with the locked step visibly locked, and the highlighted button reopens the ceremony. - **The emoji verification ceremony matches the rest of the app.** Same type, spacing and buttons as every other sheet. Cancel is no longer painted red — backing out of a ceremony is a safe, ordinary act — and the success shield uses the app's own green. - **Security warnings look like warnings again.** The "only use a code you made yourself" and "treat this like your password" callouts had been muted to grey, quieter than the buttons next to them; they carry the warning color now. Destructive actions — removing a device, stopping sync — carry an icon as well as the red, so they never rely on color alone. - **Desktop stopped stretching phone buttons.** "Add device" and the paste and connect actions size themselves to their labels instead of spanning the entire pane, and the devices panel keeps a readable width on wide windows. The sidebar's sync descriptions were rewritten to fit without trailing off mid-word. - **Planning the rest of today is faster and listens to the whole check-in.** Capture parsing and drafting now stay focused on the one artifact each step must produce, stop as soon as that artifact is saved, and immediately pick up a capture that arrived while background processing was already running. This removes avoidable AI round trips that could make planning take about a minute, while preserving separate time constraints, breaks and explicit overcommit trade-offs from dense spoken plans. - **The app has one look now.** Every screen renders the design system's own theme — the same surfaces, type and accents the newer pages were built against — instead of a third-party color scheme with the design tokens bolted on top. Backgrounds, cards and accents finally agree across old and new screens, and the brief near-black flash while the app starts is gone. - **The theme-scheme picker is gone.** Theming settings kept a list of 22 color schemes, but as screens moved onto the design system the picker changed less and less of what you actually saw — on recent screens, nothing at all. What remains is the choice that still does something: light, dark, or follow the system, synced across devices as before. - **Pairing a device is a guided flow now, not a form.** A three-station progress track — Get code, Check, Connect — draws where you are instead of narrating it, and every screen leads with the one thing to do there. On phones the camera opens straight away, with accent brackets framing where the code should land, and pasting is the fallback rather than the other way round. The check step makes the comparison the whole screen: the code your other device shows, large, with "Same code on both screens?" beneath it and the mismatch consequence riding on the decline button. Waiting is narrated by the journey's own figure — a phone and a laptop with dots streaming between them — and when the emoji match succeeds, the gap between the two machines closes into a solid line with a small celebration: your devices trust each other now, and the screen says exactly that. - **Nothing in the flow waits forever without saying what to do.** If the emoji check never appears, the last screen now tells you where to start it by hand instead of spinning indefinitely. If the camera cannot start, there is a way to try again once you have granted access. And if the device list cannot be reached while you are waiting for a new device to appear, it says so and offers a retry rather than pretending to still be looking. - **The device list reads constructively.** The roster now says how many devices are on which server, lays the cards out in two columns where the window affords it, and anchors each card on a device tile with its trust badges beneath the name. "Add device" is the page's one highlighted action, removing a healthy device is a quiet icon in its card's corner, and the stop-syncing and technical-details actions sit below their own divider at the very end. "Delete device" is now "Remove from sync", which is what it does, and a removal that the server refuses explains the situation instead of showing you the raw error. - **Setting up your very first device has a real path.** The "first device?" note used to end in "see the manual" with nothing to press; it now carries an Open the first-device guide button that lands directly on a new manual page — available in every manual language — explaining where the very first pairing code comes from and what to do with it. - **The hand-off no longer dies with the Add device sheet.** If a new device joins and finishes its emoji ceremony while you are looking at the device list — say you closed the sheet too early — the list offers Send settings and Send message history right there, instead of leaving you to find the same transfers in Maintenance. ### Changed - **The "More" menu at the bottom of the phone screen is no longer cramped.** Its rows were sized to the bare minimum a finger needs, so the destinations stacked up against each other with the icons almost filling each row. Each row now has room around it, and the list keeps an even rhythm from top to bottom. - **The linked-tasks card is a third shorter.** The relationship headings ("Follows up on", "Has follow-up", …) floated in the middle of a large empty band instead of sitting with the tasks they label, and every linked task sat in a row a third taller than it needed to be, so a card with two links took up as much room as one with five. Each heading now sits close above its own rows, the rows themselves are tighter, and nothing moves when you switch the card into Manage mode. ### Fixed - **Transcription now counts towards your AI usage figures.** When a recording was transcribed by a Melious speech model, the cost and the energy, carbon and water it used were reported by the provider and then thrown away — so the AI usage charts showed transcription as free, and every total that included it was short. Those figures are now recorded like any other AI call. - **The app no longer asks for the microphone the moment it starts.** On a fresh install, Lotti asked to record audio while you were still looking at your task list, before you had asked it to record anything — which reasonably raises the question of what it is listening to. The app was checking whether it already had the permission, purely to write the answer to its log, and on Android and iOS that check *is* the request. The check is gone: the microphone is only ever asked for when you start a recording. - **The sync statistics panel no longer shows counters that never move.** Most of what it displayed — processed messages, failures, catch-up batches, retries, circuit breaker trips and the signal counters — had stopped being recorded when sync was rebuilt, so they sat at zero forever and made a healthy sync look idle. Those rows are gone. The headline figures are now what was written to the database, how many conflicts were logged, and how much is still waiting to be applied. - **A photo or recording that never arrived now repairs itself.** If a device ended up holding an entry whose image or audio file it never received, that file stayed missing forever — nothing noticed, and only re-sending your whole history could fix it. Devices now ask each other for the specific files they are missing, and any device that has one sends it back. No action needed, and nothing is asked for repeatedly if no device still has the file. - **Photos and audio now reach a device you sync your history to.** Sending your existing entries to a newly set up device — from sync maintenance, or automatically when a device noticed it was missing entries — transferred the text of every entry but none of the images or recordings, leaving the new device with photo and audio entries it could not open. Only entries created *after* the device joined arrived complete. History transfers now carry the media with the entry. - **The bottom navigation bar no longer collides with Android's system buttons.** On phones that show the recents/home/back bar at the bottom — Samsung devices among them — the app's own bar sat partly underneath it, so its labels were crowded and taps near the bottom went to the system bar instead of the app. The bar now keeps clear of the whole system area. - **A cancelled device verification no longer claims to have succeeded.** When the other device refused or cancelled the emoji check, this one showed the green shield and "You've successfully verified…" anyway — on the outgoing screen, next to the cancellation notice at the same time. The two devices had not verified each other, and nothing said so. Cancelled and failed ceremonies now say they were cancelled, and only a real success shows the shield. - **Leaving an AI setting takes one back tap, not two.** After opening a provider, model or profile from AI Settings, backing out of the list you returned to appeared to do nothing: the list slid away and an identical list slid straight back in, and only a second tap actually left. Coming back from a detail page had quietly left the app in a halfway state that still showed the list, so the next tap was spent leaving that rather than going anywhere. Back now moves one level per tap, and the Inference Profiles page animates like a back gesture instead of like opening a new page. - **An over-full day no longer wastes a planning attempt on an impossible placeholder.** When some work did not fit, the planner could represent the omitted task as a zero-minute "unscheduled" block. The app correctly rejected that block, but only after spending the first model response. The planner is now told to name omitted work in its reasons or status note instead. - **A plan no longer treats work of unknown length as free.** When deciding what fits in a day, the planner is now shown how long each task you approved is expected to take — it previously had to find that elsewhere, and on many plans could not see it at all, so it packed in more than the day held. Tasks with no estimate are marked as unsized rather than counted as taking no time. - **Saying "nothing to do today" no longer fails.** When you spoke a note with nothing actionable in it, the assistant worked that out correctly and then had no way to say so — the app rejected the answer, and it had to either invent a task or try again. It can now record that a note held nothing, which also clears anything it had wrongly picked out of that note earlier. - **Sync finishes catching up after a large or still-growing backlog.** A reconnect could spend its recovery allowance following the newest messages, report that partial walk as complete, and leave the remaining entries missing with nothing queued to retry. Recovery now budgets the actual network work and automatically resumes whenever that allowance is reached. It also remembers the oldest unresolved encrypted message across restarts, so a newer applied message cannot move recovery past a gap that still needs its decryption key. - **The planner is told how much of the day is actually left.** It knew your capacity and your working hours, but had to work out for itself how much time a mid-afternoon plan really had — and got it wrong, running plans past the end of the working day. It is now given the remaining minutes directly, and when the work on offer does not fit it is asked to say so: leave something out and name it, or schedule part of a task and say how much is left, rather than quietly shrinking everything until it appears to fit. - **Time on a task is credited to that task's area.** A planned block could name a task from one area while being filed under another, so the day's per-area totals — and the weekly rollups built on them — quietly credited the wrong one. A block that names a task now always follows that task's area. - **The planner can no longer mark blocks as agreed on your behalf.** A fresh plan could come back with blocks already marked committed — the state that means *you approved this* — so work you had never seen presented itself as settled, and uncommitting the day was the only way to undo it. Only you can commit a block now. A re-plan still carries forward blocks you already committed, unchanged. - **An entry's time follows your device's clock.** If your device is set to a 24-hour clock, an entry recorded at 21:54 said "9:54 PM" at the top of the entry — while the editor that opens when you tap it correctly said 21:54. The header only ever asked the app's language what a time looks like, never the device, so an English-language app was locked to AM/PM whatever the device was set to. Both now agree, in every language. - **The entry header no longer has a hole in the middle of it.** The category button sat between the date and the rest of the buttons but belonged to neither, so the leftover width in the row collected on one side of it — a gap that grew with the window and, on a wide screen, left the category stranded in open space. Date on the left, every button together on the right, evenly spaced. - **The date at the top of an entry is no longer shouting.** It sat at the same size as the app's own title text, and heavier than the body text beneath it, which made the timestamp the loudest thing on a card whose actual content sits underneath. It now sits one step down, on every screen — including the narrow phone layout that splits the date over two lines. - **Times in your day plan follow your device's clock.** If your device is set to a 24-hour clock, a block running 14:30–16:05 still read "2:30 PM–4:05 PM" on the agenda card and in the block editor. The app only ever asked its own language what a time looks like, never the device, so an English-language app was locked to AM/PM whatever the device was set to. It now asks both. - **A block shows the same time wherever you look at it.** The plan list in Refine used a different clock from the rest of Daily OS, so one and the same block could read "2:30 PM" on the agenda and "14:30" while you were talking about changing it — on every device, in every language. Both now agree. - **An event's times follow your device's clock too.** The date under an event title, every time down its timeline, and the end of a recorded session were all fixed to a 24-hour clock, so an event at half past two read "14:30" even on a device set to AM/PM. They now read the way the rest of your device does. - **Chart tooltips are readable in your language and on your clock.** The date in a chart's tooltip was always in English — "Mar 15" whatever language the app was in — and its time was always 24-hour. Both now follow the app's language and the device's clock setting. - **The blocks on your day's timeline read on your clock.** Every block showed its span as "14:30–16:05" regardless of the device — and the proposed change describing that same block used a 12-hour clock instead, so the two disagreed. Blocks and proposals now both follow your device. The hour scale down the side of the timeline stays 24-hour: it is a ruler, and its end-of-day mark has no unambiguous 12-hour spelling. - **Switching Time Analysis to Day shows today.** Choosing Day took you to the first day of whatever you were looking at — January 1st from the year view, the 1st of the month from a month — so the page you landed on was almost never the one you wanted. It now opens on today whenever the period you are in still includes it. Looking at a past period still keeps your place there rather than jumping you back to today. - **The date picker in Time Analysis no longer changes size as you browse.** Months need five or six rows of dates depending on their length and which weekday they start on, and the card grew and shrank to match — so paging through months moved the arrows out from under the pointer. The card now reserves room for the longest month and keeps one height throughout. ## [0.9.1070] ### Fixed - **Removing a dead device now unblocks sync immediately.** Deleting a device from the sync device list refreshes the cached device keys and nudges the pipeline right away, so the remaining devices resume syncing without an app restart. Deleting also asks for confirmation first — it signs the device out of the sync account — and a rejected password now gets a clear explanation instead of a raw error dump. ### Added - **Decide per area whether assistants keep themselves up to date.** New tasks in an area got an assistant that only ever updated when you asked it to, and turning that on was a per-task chore. Areas with an agent template now carry a "Wake the assistant automatically" switch that decides what new tasks start with. It's separate from the area's automatic-AI switch on purpose — transcription and image analysis keep running whether or not assistants update themselves. The switch applies to tasks created after you flip it; tasks you already have keep whatever their own switch says, so nothing you set by hand gets overwritten. ### Changed - **The AI summary card's update controls read as two clear groups.** "Is this current, and can I refresh it now" sits on the left — a freshness word next to the button that acts on it — and "does it update itself, and when next" sits on the right, where the countdown now belongs to the switch it comes from instead of being stranded on the other side of the card. The manual button is also deliberately quieter than "Confirm all" above it, so the loudest thing on the card is the one that changes your task, not the one that spends tokens. - **Cancelling one scheduled update is now "Skip once"**, so it is clear it skips a single run rather than turning automatic updates off. - **Model names shorten by dropping whole parts instead of cutting words.** A route that no longer fits now goes "Qwen 3.5 Plus · Alibaba · via Melious.ai" → "Qwen 3.5 Plus · Melious.ai" → "Qwen 3.5 Plus", so the model and the provider stay readable instead of trailing off mid-name. ### Fixed - **Sync no longer loses messages that arrive before their decryption key.** When a lot of entries arrived at once over a slow connection, some could be discarded while the app was still waiting for the key to decrypt them — leaving that device permanently short of entries the other device had, with nothing pending and no way to notice. Entries now wait for their key for a full ten minutes regardless of how busy the device is. - **A screen reader can now activate the whole "Automatic updates" row.** Tapping the wording flipped the switch for everyone else, but assistive technology could only reach the switch itself. - **Planning a day that has already started no longer wastes a round trip.** When you asked for a plan mid-afternoon, the planner would sensibly start your first block "now" — and the app rejected the whole plan for starting in the past, because by the time the plan came back, "now" had moved on. The planner had no way to win: it would have had to predict how long it was going to think. It is now told a start time that is still valid when the plan lands, so the first attempt succeeds instead of the second. - **The update section at the bottom of the AI summary card takes up much less room, especially on a phone.** The switch row in particular had far more space above and below it than it needed, which pushed the model line down and made the settings area compete with the summary itself. The card is now about 36 points shorter on a phone, and 69 shorter in German at large text sizes. - **Event and project assistants show up right away too.** Creating an event or a project with an assistant left its card empty for several seconds before the assistant appeared — the same delay tasks had, from the same cause. It now appears as the page opens. - **The update section at the bottom of the AI summary card now reads as its own band in dark mode.** Its tint was mixed the wrong way for a dark background, so it was all but invisible against the card; it now separates the way it always did in light mode. - **Day plans stop scheduling work that is waiting on something else.** When the planner drafted a day without a voice note to work from — a scheduled draft, or the "plan my day" button on its own — it was told to respect blocked tasks but never shown which tasks were blocked, so it would place blocked work as if it were ready. Blocked tasks now arrive with their status and what they are waiting on, and the planner either schedules the blocker first or leaves the blocked task out and says why. - **The bottom of the card lines up.** Every row in the update section now starts on the same edge, including on a phone and in longer languages, and the card keeps the same bottom margin whether or not the "This report …" line is showing. - **Choosing your own model no longer switches off automatic transcription and image analysis.** Picking a model by hand for a task's assistant replaced the whole AI setup, not just the thinking model, so the category's automatic transcription and image analysis quietly stopped — and picking a different model afterwards never brought them back. Each of them now falls back to the profile the task inherits from its category, so they keep using the category's transcription and image models while your chosen model handles the thinking. A profile you picked for a task still wins every capability it actually configures; only a missing one falls through. This also fixes tasks created before their category had an AI profile, which never picked one up. - **A new task gets its assistant right away.** Creating a task in a category with an agent template left the AI card empty for several seconds before the assistant appeared, which looked like assignment had failed and led people to add one by hand. The assistant was in fact created immediately — the card just had no way of hearing about it until the first update finished running. It now appears as the task opens. ## [0.9.1069] ### Fixed - **Ticking off a checklist item no longer looks broken as the row leaves.** With the Open filter on, a checked item's row shrank away but its checkbox stayed full size and got sliced in half by the shrinking row, while the rest of the row jumped sideways the instant the animation began. The row now shrinks as a single piece — checkbox, text and buttons together — and fades out smoothly. - **Your day plan can no longer contain tasks from categories the day agent isn't allowed to see.** The list of tasks a plan was permitted to reference was taken from the agent's own request without being checked, so a task from any category — or one that had been deleted, or never existed — could be attached to a block just by naming it. Task references are now resolved against your journal and filtered by category, the same way every other task reference already was. - **No more phantom calendar entries in your day.** Day plans could include blocks marked as imported calendar events, which the app then refused to let you edit — telling you to change them in the source calendar, where they did not exist either. The day agent is never shown your calendar, so it can no longer claim an event came from one. - **Nothing gets planned into the part of today that has already passed.** A block could still be placed in the past by marking it as already committed; only genuinely started, finished, or dropped work may sit earlier than the current time. - **You can trigger an update at any time again.** While an automatic update was scheduled, the manual button disappeared entirely, so the only way to run the agent by hand was to cancel the countdown first. The button is now always there, and using it no longer drops the schedule. - **The "This report …" line no longer breaks mid-sentence.** When the summary on screen was written by a different model than the one currently selected, the attribution wrapped onto a second line and left a stray fragment behind; it now stays on one line, with the full text on hover. - **Hover highlights in the AI summary card no longer stretch across the whole card.** Pointing at the model row or the card header highlighted the entire width instead of the thing under the pointer. - **A soul's history no longer blinks while it updates.** On a soul's version history and evolution history, any change to that soul — an edit, a finished ritual session, a sync arriving — swapped the list you were reading for a loading spinner and back, collapsing the section and shifting the rest of the page. Those lists now stay on screen while they update. The same applies to an agent template's reports list. - **Answering a task suggestion no longer jumps the page.** Checklist suggestions already stayed put, but the others did not. Accepting or dismissing a due-date, language, priority, label or estimate suggestion while you had scrolled past the assistant card yanked whatever you were reading upwards, because the answered row's card shrinks as it collapses and nothing was holding the content below it in place. Accepting a follow-up task moved the page too, and could move it again seconds later when the new task finally appeared in the linked-tasks list. All of these now hold their position, and scrolling still cancels the hold immediately so nothing ever fights you. - **Deleting a bundled AI profile or model now sticks on every device.** Deletions were remembered only on the device where you made them, so a second device could re-create the profile and sync it back. The deletion now travels with your other AI settings, so it holds everywhere. Removing a provider still takes its models with it and still restores them if you add that provider again, and re-running onboarding still brings back that provider's bundled profile. - **Entries you mark private are stored private.** Anything created with privacy set but without being attached to a parent entry — a private measurement, a completion of a private habit — had that flag cleared as it was written, and was stored as an ordinary entry. Privacy is now kept through the write. Entries attached to a parent still take the parent's setting, which is unchanged. ### Added - **Link a task that doesn't exist yet, without leaving the picker.** Searching for a task to link used to dead-end on "No tasks found" when the work simply hadn't been written down. Typing a title that matches nothing now offers to create it: the task is made with that title, put in the same category and project as the task you're linking from, connected with the relationship you already picked, and you stay where you were — with the usual Undo. The "Create new linked task…" menu entry still does what it always did, creating a blank task and taking you to it. ### Changed - **The AI summary card's bottom section is simpler, and works on a phone.** The update controls used to sit in a box inside a box; that inner box is gone, so the whole strip reads as one quiet band and no longer runs out of room on narrow screens or in longer languages. As space gets tight the countdown shortens its wording — "Next update in 1:30", then "in 1:30", then just "1:30" — instead of cutting off the time itself, and the controls stack in a readable order rather than being flung to opposite edges. - **"Wake agent" is now "Update now",** and the summary says "Up to date" or "Out of date" in words rather than only as a coloured icon. - **Cancelling a scheduled update is now a labelled "Skip"** instead of an unmarked "✕" next to the Automatic updates switch. It still cancels only the pending run and leaves automatic updates on. ## [0.9.1068] ### Changed - **Dropdowns and search boxes now look like the same control.** Where the two sit stacked — choosing how a task relates, then searching for the task to link — the dropdown was drawn as a flat opaque box with much rounder corners and a heavier outline than the search field, so the pair read as parts of two different apps. The dropdown now matches the search field's shell, including in dark mode, where it lifts off the panel it sits on instead of looking like a hole punched in it. - **Warnings, errors and status text are readable in light mode.** The amber, green and blue used for warnings, successes and information were mixed too light to stand out against a white or off-white background — amber measured barely above the page it sat on, so a "blocked" mark or a pending-sync notice could all but disappear. Those three tones are now deep enough to read, and any text carrying one of them — priority labels, status chips, connection results, ledger figures — is picked to stay legible in whichever theme you use. Dark mode is unchanged apart from error text, which lightens slightly on cards. - **Areas you create during onboarding start with automatic AI on.** Automatic transcription and image analysis remain off by default for every category you make later — but the onboarding flow, where you connect a provider and record your first capture on purpose, now switches it on for the areas it creates, so the app keeps working the way onboarding just showed you. If you reuse an area where you had already switched automation off, it stays off. ### Fixed - **Re-running onboarding restores a bundled profile you had deleted.** Setting up a provider through onboarding again now brings back that provider's bundled profile even if you had removed it earlier, so the areas onboarding creates always point at a profile that exists. Deleting it stays permanent everywhere else. ## [0.9.1067] ### Fixed - **Deleting a bundled AI profile or model now sticks.** The app re-created its bundled defaults whenever one was missing — at every launch, and again each time you saved a provider — so deleting one you did not want brought it straight back, sometimes within the same session. Deletions are now remembered. Removing a provider still takes its models with it and still restores them if you add that provider again. - **Clearing a profile's automated skills no longer re-enables them.** If you removed every skill from a bundled profile, the next launch put them back with automation switched on. Your empty list is now left alone. ## [0.9.1066] ### Added - **You now hear about a day plan that failed, not just one that worked.** If drafting or refining a plan gave up while Lotti was in the background, nothing told you — you came back to a day that had simply never been planned, with no explanation. Those failures now raise a notification that takes you to the day so you can retry. ### Changed - **The "Active" badge on inference profiles now means something you can control.** A profile was badged Active purely because one of its models belonged to a provider you had connected — so a profile could appear active the moment you set up a provider, with nothing you could do to turn it on or off, while the profile your categories actually used showed nothing. The badge now reads **In use** and appears only when a category or a task agent actually routes through that profile. - **A provider's page now lists every profile that depends on it**, under "Profiles using this provider", instead of naming a single arbitrary "Active profile" — so you can see what would break before removing a provider. ## [0.9.1065] ### Changed - **Automatic AI now waits to be asked, per category.** Transcribing new audio and analyzing new images automatically is now an explicit per-category switch, sitting under the category's AI profile. Previously, picking a profile was enough to start it — profiles arrive with automation already enabled — and audio could even be transcribed automatically with no profile chosen at all. **The switch starts off everywhere, including categories that already had a profile**, so turn it on for each category where you want AI to keep working in the background. Per-recording "don't transcribe this one" still works as before. ### Added - **Turning on automatic updates now wakes the task agent right away.** The switch used to only arm future updates, so you had to tap "wake agent" yourself to get a current summary. It now fetches one immediately — unless the summary is already up to date, in which case it stays quiet rather than spending tokens on a refresh you don't need. ## [0.9.1064] ### Added - **Undo a link you just made.** Picking a task in the link or blocker picker still commits on that single tap — but it now says what it wrote ("Blocks: Draft crew rotation schedule") and offers Undo, so a mis-tap or the right task under the wrong relationship is one tap to take back instead of a trip through Manage links. ### Fixed - **The Linked Tasks card cut off its own name** in German and Spanish, and at larger text sizes in every language. The Link action now drops to its icon where the label won't fit beside the heading, rather than the heading giving way. - **The unlink confirmation names the task** you're removing, instead of asking about "this task" from a row of unlabelled icons. - **The Linked Tasks header broke at larger text sizes.** At accessibility text scales the Link button was cut off and the ⋯ menu vanished entirely — which, while editing links, was one of only two ways out. - **Undo after linking left the message on screen.** It removed the link but kept showing the confirmation and a live Undo button for another ten seconds, offering to undo something already undone. - The confirmation now shows the linked task's full name instead of cutting it off, and modal titles shorten with an ellipsis rather than reflowing their bar in longer languages. - **A task you had already linked could be linked again.** A plain link reads the same from either end, but the picker only recognised it from the side it was created on — so a task already listed on the card was offered again, and taking it left two identical rows and needed two confirmations to remove one link. - **Failures said the wrong thing and looked like successes.** Every rejected link reported "this would create a blocking cycle", even for relationships that can't create one, and failure messages arrived on the same background as the "link created" confirmation. - Manage mode has a way to add a link again — its header slot is taken by the Done button while you're editing. - **The link and blocker pickers listed no tasks at all.** Browsing (before typing anything) came back empty in every case, because the query treated "all categories" as "no categories". Searching still worked, which is why it looked like a search quirk rather than an empty list. - **A task with links briefly claimed it had none.** Opening a task showed the "Link a task…" empty prompt until its links finished loading, so every populated Linked Tasks card flashed an empty one first. - **Finished tasks can be linked again.** "Follows up on", "Duplicates", "Fixes" and "Supersedes" all commonly point at work that is already done, but completed tasks were excluded from the picker and it reported "No tasks found" for a task you could see everywhere else. Blockers still list only open tasks — a finished task can't block anything. - **Undo after linking now tells you when it fails**, instead of silently leaving the link in place, and it stays on screen long enough to reach. - Edit and unlink in manage mode meet the 48pt minimum touch target; they were 40pt, side by side, over a row that is itself tappable. - **Task search could deny a task that exists.** In the link and blocker pickers, full-text search results were matched against only the first 200 open tasks loaded, so on a larger backlog a task you searched for by name could come back as "No tasks found". Matches outside that window are now fetched directly. ### Changed - **An empty Linked Tasks card offers both actions in words.** Creating a new linked task used to live only behind the ⋯ menu, on the one screen where you have nothing to go on yet. - **Linking a task is a labelled button again.** Once the Linked Tasks card had any links, its only way to add another was an unlabelled icon; it now reads "Link", matching the Done button that shares that slot. - **Save in the edit-relationship sheet stays off until you change something**, instead of offering to save an edit nobody made. - French and Romanian task statuses are sentence case ("En cours", "În progres") rather than all caps, matching every other language. - **One word for one relationship.** The header chip on a blocked task reads "Blocked by 2 tasks" again, matching the section it jumps to and the option that created the link — previously it said "Waiting on", so the same relationship went by two names depending on where you met it. - **The edit-relationship sheet names the task you are editing the link to.** It read "This task… / Is blocked by" with nothing after it, leaving you to remember which row you had tapped. - Editing links no longer hides each linked task's status — knowing a blocker is already done is most useful exactly when you are deciding what to unlink. - **Manage mode says it is on.** The Linked Tasks card now shows a Done button in its header while you are editing links, instead of leaving the only exit inside the same overflow menu you opened it from. - On a desktop-sized window the task search field is focused when the link and blocker pickers open, so you can type straight away. Left unfocused on phones, where it would raise the keyboard over the results. - The Linked Tasks overflow menu uses the app's own surface and radius rather than Material defaults, so it matches the card it opens from. ## [0.9.1063] ### Changed - **Picking what two tasks mean to each other is now one choice, not two.** The relationship picker was a row of type chips plus a separate direction switch, which put the same word on screen twice — a chip reading "Blocks" directly above a switch whose first option also read "Blocks". You now pick from a single list that completes the sentence "This task…": Relates to, Blocks, Is blocked by, Follows up on, Has follow-up, Duplicates, Is duplicated by, Fixes, Is fixed by, Supersedes, Is superseded by. - **You can change a relationship after creating it.** Turn a plain link into a blocker, or reverse which of the two tasks is the blocker, without unlinking and starting over. - **The Linked Tasks card groups links under the exact phrase you picked**, so it reads back the words you chose. What the task is waiting on comes first and is the one thing called out; everything else stays quiet. Each row shows the linked task's status. - **The "waiting on" chip in the task header is compact.** It states that the task is waiting and how many tasks it waits on; scroll to Linked Tasks to see which. Previously it embedded the blocker's full title and could span the whole header. - **Linking, blocking and editing sheets now adapt to the window** — a bottom sheet on a phone, a centred dialog on a wide window, matching every other task picker — and size themselves to their content instead of leaving a large empty area below a short list. ### Fixed - **Two tasks can hold more than one relationship.** A task already linked to another could not be linked again with a different relationship — the second one silently reported "No tasks found". - **A task with no links can now be linked.** The Linked Tasks section was hidden entirely until a task already had a link, and the only way to add one lived inside that hidden section. It now always offers a "Link a task…" action. - Long task titles no longer truncate mid-word when picking a task to link. ## [0.9.1062] ### Added - **AI activity now uses one consistent decoder-bars animation.** Unified AI progress, entry details, and cover-art generation now match the shader-based activity indicator already used by task details and Daily OS, including smooth entry and exit transitions and tap-to-open progress where available. - **Task agents can now read your image analyses.** The AI analyses attached to a task's images — brief summaries and full OCR extractions alike — are now part of the task agent's context, each with the model that produced it, when it ran, and its full text. The agent can finally act on what's *in* an image: extract an appointment date from a photographed card and propose it as the due date, name the task after the document, or reference OCR'd details in its report. A completed analysis now also flags the task for the agent's next regular update — previously the agent was never told at all. - **AI analysis cards are collapsible and visually calmer.** Long analyses (e.g. a full OCR extraction) under an image or audio entry now start fully collapsed — just a "Show more" toggle and the model pill — so one card can no longer dominate the entry; "Show less" folds it back down. The cards also drop their heavy card-on-card look for the app's tinted AI surface — the same color family as the agent report cards, with a soft accent hairline, in both light and dark themes. - **Task links now carry real relationships, not just a plain connection.** When linking two tasks you can now pick what they mean to each other — Blocks, Follows up, Duplicates, Fixes, or Supersedes — in either direction, both when linking an existing task and when creating a new linked task. The Linked Tasks section groups these by relationship instead of one flat list, and a task that's blocked shows a "Blocked by" chip right next to its status in the header — tap it to jump to the blocker, or see all of them when there's more than one. Marking a task Blocked now optionally prompts you to name what's blocking it on the spot. - **"Your plan is ready" as a system notification.** When a Daily OS plan draft or refinement finishes while the app is in the background, Lotti now raises a notification instead of completing silently — tap it to jump back to your day. ### Fixed - **Photo rows in the Logbook no longer shout.** An image entry's caption was rendered several sizes larger than the title of every other row in the same feed, so photos dominated the list. It now sits on the same title scale as note, task and event rows, and the timestamp beneath it matches the other cards' brightness. - **Day-plan processing can no longer retry forever in the background.** A draft or refinement whose AI call completed without producing the expected result used to re-run a full model request every few seconds, indefinitely and invisibly. Such failures now count toward the retry cap and stop with a visible failure after a few attempts. A refinement waiting for a plan that will never be drafted also fails fast instead of spinning forever. - **A committed day plan can no longer be silently replaced.** The planning agent could overwrite a plan you had already committed — including block progress — with a fresh draft, bypassing the change-approval flow. It is now told to propose changes for your approval instead. - **Two quick plan refinements no longer swallow the second one.** Asking for a second change while the first was still processing could mark the second request as done using the first one's proposed changes, silently dropping your instruction. Results are now matched to the exact request that produced them. - **Rebuilding your day after a crash uses your latest selections.** If the app was killed mid-draft, tapping "Build my day" with new decisions used to silently run the old ones; the stale attempt is now replaced. - **The day's Activity timeline reflects processing state live.** Cards could freeze on "Transcribing…" or miss retry/waiting-for-network/failure transitions until something unrelated refreshed the screen; every processing state change now updates the timeline, and cancelling a recording's processing is reflected immediately. - **Two devices working on the same day no longer strand its plan.** When one device created the day's dedicated agent before another device's work on that day had synced in, the day's plan and pending changes could become permanently invisible and a check-in could get stuck retrying its parsing forever. All day reads now accept both legitimate writers of a day. ## [0.9.1061] ### Added - **The Daily OS coordinator now briefs each day's planning agent.** Every morning a coordinator digest reviews escalations and recent days, then issues a distilled directive per day — your commitments, the day's capacity budget, carry-overs, and cross-day context — which that day's agent must honor when drafting: every commitment is placed, explicitly traded away with the collision named, or escalated back. Day agents can raise typed status events (over-committed, directive unsatisfiable, plans diverging, processing blocked) that the coordinator picks up at its next digest, so week-level plans and learnings persist while each day keeps a small, fast context of its own. Each digest also pools the last four complete weeks into per-category planned-vs-recorded rollups, so the coordinator can steer month-scale pacing — a chronically shortchanged category or a shrinking planning streak — not just the day at hand. ### Fixed - **The green focus border in list boxes no longer gets cropped at the corners.** In the task filter box (and every other grouped list, e.g. in Settings), the first and last row's keyboard-focus highlight was drawn as a sharp rectangle and clipped by the box's rounded corners — visible after picking a Status or Project filter. Edge rows now round the highlight to match the box's corner radius. ## [0.9.1060] ### Added - **The Manual's five walkthrough videos are now available in every language Lotti supports** (previously English and German only): French, Italian, Spanish, Czech, Dutch, Romanian, Portuguese, Danish, and Swedish. - **Manual videos now have a mobile-view variant, matching the existing screenshot toggle.** The same mobile/desktop switch that already controlled screenshots now also switches the walkthrough videos, and first-time visits from a phone or narrow window default both to mobile view automatically; saved viewport preferences continue to take precedence. - **Voice captures now survive the app closing before they're parsed.** Submitting or retrying a Daily OS capture queues its parsing durably (the same restart-safe queue drafting already uses), so backgrounding or killing the app between recording and parsing no longer loses the parse — it runs when the app reopens. ### Fixed - **The Settings list no longer jams its last entry (the Manual) against the bottom navigation bar on phones.** The list now scrolls clear of the docked nav pill, so the Manual row is fully readable and tappable. - **The Keyboard Shortcuts page now smoothly hides the bottom navigation bar**, matching the slide-away motion every other terminal settings page (Daily OS, Theming, Speech, …) already had. - **The Projects list's floating "new project" button no longer covers the last project card.** The list reserves clearance for the button and the nav bar — including the home-indicator inset on modern phones — instead of a fixed spacing that came up short. - **Opening a project from the Projects tab now slides the bottom navigation bar away**, giving the project detail page the whole bottom edge — the same motion as the settings detail pages. - **The category icon picker no longer overflows its grid tiles on narrow phones.** On ~375pt screens the icon label could spill out of its tile by a few pixels; the label now shrinks into the available space instead. - **Same-day plans can no longer schedule new blocks in the past.** The drafting guard that kept AI and manual blocks from starting before the current time now covers buffer blocks too; only imported calendar events — which legitimately span the current moment — are exempt. - **The Settings destination is readable again in Italian and Swedish.** Several Settings-related labels (the main navigation destination, agent editor tabs, the logging screen, and the Settings breadcrumb) showed "Impostazioni delle impostazioni" in Italian and "Miljöer" in Swedish instead of "Settings" — now "Impostazioni" and "Inställningar". - **The coding-prompt "Copy Prompt" button is correctly labeled in Dutch.** It previously read "Waarschuwing kopiëren" ("Copy warning"); now "Prompt kopiëren". ## [0.9.1059] ### Added - **Five narrated walkthrough videos are now embedded in the Manual, in English and German.** Creating a task with your voice, generating cover art, turning a coding task into an AI-ready prompt, filtering and saving task views, and setting up a category each have a short captioned video on their Manual page, alongside the existing screenshots. ## [0.9.1058] ### Fixed - **Accepting AI suggestions no longer bounces the task page.** Confirming a batch that mixes header changes (labels, due date, priority, status, title) with checklist items used to shove the page down when the header grew and then snap it back, leaving the rest of the batch uncorrected. The task header now reports its height changes to the same pre-paint scroll stabilizer as the checklist, and the page's two stabilization layers now recognize each other's corrections instead of mistaking them for user scrolls and disarming mid-batch. ## [0.9.1057] ### Added - **Drafting or refining a day's plan now survives the app closing mid-request.** The request is queued durably instead of held only in memory, so if the app is backgrounded or killed while a plan is being drafted or refined, the work finishes in the background and the result is there when you come back — instead of the request silently vanishing. - **Each day now gets its own planning agent.** The Daily OS assistant used to run every day's planning through one shared identity; going forward, a new day gets a dedicated agent the first time you use it, so its wake history, token spend, and conversation log (Settings > Agents > Inspect agent) show that day on its own instead of mixed into one long-running timeline. Past days you've already planned are unaffected. ### Removed - **The agent Stats tab is gone.** Settings > Agents now opens on Templates and keeps four views: Templates, Instances, Souls, and Wake Cycles. The token-usage dashboard (daily/per-model charts, per-agent breakdown, and 24-hour wake histogram) has been removed along with its sidebar entry — AI usage reporting lives in the AI Usage view instead. ### Added - **"Here's what I heard" is now visibly busy while it works.** The thinking shader animates on the action bar's top edge — the same signal as the other planning steps — until parsing fills the Heard column. Previously the screen sat frozen for many seconds because the busy indicator was wired to a signal that never fired. - **Edit a day recording's text right on its Activity card.** The separate text dialog is gone — every unsubmitted recording now embeds the same rich-text editor used across the journal, with its familiar toolbar and save shortcut. Text you write or correct there is treated as your final wording: a pending transcription completes instantly with your text, and a transcript that arrives later is kept as a record but never overwrites what you wrote. - **AI work now shows its real footprint where it was created.** Coding prompts, generated and analyzed images, and audio transcripts carry a tappable attribution pill. Its details show the creator, model, request evidence, available input/output/cached/reasoning tokens, and the provider's actual cost. When Melious reports it, the same view also shows energy, CO₂e, and water use. Audio entries show the total across all their transcripts, while each transcript retains its own attribution. Coding-prompt usage also contributes to its parent task's AI-consumption total. ### Changed - **Generated cover art no longer clutters the task's timeline.** The banner and list thumbnail are unchanged, but the generated image's own entry in the task's linked entries now starts collapsed instead of showing the same artwork a third time — tap it to expand if you want to see it there too. - **The task AI summary card's wake control no longer moves around.** Freshness (out of date / up to date), the wake button or countdown, and the automatic-updates switch now always live together in one row, instead of the wake button appearing in a separate strip above the switch only while automatic updates were off. Model identity is now always its own row below, never squeezed beside that cluster. ### Fixed - **Generate Cover Art now actually shows the generated image.** The task was pointed at the wrong internal id after generation, so the artwork was created successfully but never appeared in the task header or task list. - **The record/stop button no longer flashes both states at once.** Starting or stopping a recording briefly showed the record button overlapping the stop/pause/cancel row mid-transition. ## [0.9.1056] ### Added - **A recording of your day can no longer get lost.** Stopping a Daily OS check-in now saves the audio to your journal immediately — before any transcription is attempted. If you are offline or the transcription service fails, the recording stays safe and is retried automatically in the background once you are back online. - **New Activity view for your day.** The Day header now offers Agenda | Day | Activity. Activity shows everything that happened around your day in order: saved recordings (with playback), check-ins, the generated plan, and day summaries. Recordings waiting for transcription show their status with retry and "add text yourself" actions, and a transcript can be sent straight into planning or refining. - **The planner sees your recordings.** Day recordings become visible to the day agent as soon as their transcript exists — even if you never manually submitted them — so a later planning session picks up what you said while offline. ### Removed - **Live (realtime) transcription is gone.** The realtime mode toggle in the recording sheet and chat input has been removed. All transcription now runs over the finished recording, which allows better provider choice and speech dictionary support; the transcript appears a few seconds after you stop instead of streaming word-by-word. ### Changed - **The logbook now matches the rest of the app.** On desktop it uses the same split layout as Tasks: the entry list stays on the left, and opening an entry fills a reading panel on the right instead of navigating away — with its own resizable divider, a highlighted selected row, and no duplicate back button or add button in the panel. The list got the same titled header as Tasks and Projects (Logbook title, compact search, filter), denser cards with the same font sizes as task rows, and the entry details page now shares one unified background with the list, presenting the entry as a card with text capped at a comfortable reading width on wide windows. On phones the logbook still opens entries as their own page, restyled to match. ## [0.9.1055] ### Fixed - **Lotti now closes cleanly on Linux and macOS.** Desktop shutdown uses one ordered teardown that stops background services and releases every SQLite database before the Flutter engine exits, avoiding the SIGABRT crash seen after closing the app. - **Secure storage works again in the Flatpak build.** The sandbox can reach the desktop Secret Service, so Matrix credentials and other protected values can be read and persisted instead of failing with a misleading locked-keyring warning. ## [0.9.1054] ### Added - **AI-generated work now carries a consistent attribution record.** Coding prompts, generated and analyzed images, audio transcripts, and agent reports show who or what initiated the work and open a shared detail view for model, usage, timing, interaction evidence, and provider-reported cost. Attribution is preserved across encrypted sync. Existing work without attribution stays unattributed rather than receiving a guessed creator or cost. ### Fixed - **Task AI summaries are cleaner, more compact, and easier to operate.** Report and proposal text now match editor entries and card summaries, quiet metadata is easier to read, secondary controls have larger interaction targets, and the footer adapts cleanly to narrow layouts and larger text. Scheduled updates and their switch now form one responsive automation group beside the model on wide cards and above it on phones; countdown ticks no longer shift neighboring controls. Canceling automation while an update is pending also keeps the summary correctly marked as out of date. Accepting task-agent suggestions no longer shifts the task detail scroll position when checklist and proposal content resize together. - **Date pickers now start the week according to your device region.** Entry dates, entry ranges, due dates, and project target dates show Monday first in European regions and Sunday first in the US, including matching weekday headers and calendar-day positions. - **Screen-reader controls now consistently expose one actionable, localized announcement.** Back, overflow, chart-baseline, and entry-collapse controls retain their labels, actions, and state without duplicate speech; new accessibility labels are available in Danish, Swedish, and Dutch. - **Matrix sync now finishes catching up after reconnecting during a large burst, even on poor networks.** Reconnected devices recover every missed entry instead of stopping after a partial server page. ## [0.9.1053] ### Added - **Danish and Swedish are now available across Lotti and its Manual.** The app follows Danish and Swedish system locales, with matching Manual routes, localized screenshots, and explicit language choices in Settings → Advanced → Language. ## [0.9.1052] ### Added - **Dutch is now available across Lotti and its Manual.** The app follows Dutch system locales, and the published Manual now has a Dutch route with matching localized screenshots. Settings → Advanced → Language can also select Dutch explicitly. ### Changed - **Language now changes Lotti as well as the Manual.** Selecting a language in Settings → Advanced → Language updates the app UI immediately and opens the same Manual translation. Follow system restores your device language for both. ### Fixed - **Screen-reader navigation, fields, and visual controls now describe their actions and state.** Overflow menus, text-field icon actions, search clear, chart-scale controls, and calendar dates expose accessible names; fields keep their labels and validation feedback when focused; back navigation follows the app language; and unavailable navigation destinations announce their disabled state instead of sounding actionable. - **The task AI summary card is more compact and calmer.** The summary now leads directly under a tighter header, and all secondary controls moved into a quiet footer at the card bottom: the Wake agent button, a small countdown chip while an automatic update is scheduled, the automatic-updates switch, and the model/provider line. While automatic updates are off, a slim status strip under the summary always reserves the same space — "This summary is out of date" with a right-sized Wake agent button, or a quiet "Summary is up to date" confirmation — so the card never jumps when freshness changes. Proposal rows now lead with the proposed change itself over quiet accept/reject icons and a neutral kind tag, long summaries keep a readable line length on wide windows, and History plus Confirm all share one row under the list. Duplicated explanatory captions are gone. ### Fixed - **Audio transcripts respect the automatic-updates switch.** Finishing an audio transcription — locally or synced from another device — no longer wakes the task agent while automatic updates are off. The summary is marked out of date instead, so the card shows the Wake agent action and no inference tokens are spent without consent. ## [0.9.1051] ### Fixed - **The entry date stays readable on narrow phones.** When an entry header's action buttons (AI assistant, flag, star, overflow menu) would leave the timestamp squeezed into an ellipsis, the action rail now uses tighter 4px gaps so the date and time remain visible while every button keeps its full-size tap target. ### Added - **Italian is now available across Lotti and its Manual.** The app follows Italian system locales, and the published Manual now has an Italian route with matching localized screenshots. Settings → Advanced → Language can also select any published Manual language explicitly. - **Portuguese is now available across Lotti and its Manual.** The app follows Portuguese system locales, and the published Manual now has a Portuguese route with matching localized screenshots. Settings → Advanced → Language can also select Portuguese explicitly. - **Romanian Manual and language-aware links are now available.** Lotti opens the published Manual in English, German, French, Czech, or Romanian based on your system language, and Settings → Advanced → Language lets you choose one explicitly. - **Manual is easier to find without crowding the sidebar.** It now sits in a distinct final section of Settings and in the desktop Help menu, opening the current published manual in your default browser. ### Changed - **Czech now reads consistently across Lotti and its manual.** Category, dashboard, habit, time-analysis, session-rating, reference-image, and sync labels use one informal Czech voice, with the accompanying manual and screenshots using the same terms. - **Absolute task due dates now use the selected app language.** The desktop task header and reusable due-date label format calendar dates in the active locale instead of falling back to English month names. - **French, Spanish, and Romanian now read more naturally throughout the app.** The selected language keeps its own register and terminology across agent controls, AI and Sync setup, dashboards, health import, and task actions, instead of exposing literal or mixed-language copy. - **Events and built-in surveys now follow the app language end to end.** Event status and date labels, the PANAS questionnaire, and its embedded navigation, progress, validation, and dismissal controls no longer fall back to English in German or Czech. - **Czech and German setup, automation, and release-note screens now stay in their selected language.** Agent controls, inference profiles, labels, Sync activity, onboarding diagnostics, maintenance previews, and the newest release badge no longer fall back to mixed English UI copy. - **German and large-text layouts stay usable in key setup flows.** AI form status text truncates without pushing its actions off-screen, Daily OS time anchors wrap at accessibility text sizes, and the What's New footer gives translated actions enough room on phones. - **Onboarding now follows the active light or dark theme throughout setup.** Panels, text, controls, provider tiles, recording previews, and task handoff states use the matching design-system tokens. The neural and aurora artwork also adapts its palette and compositing so the motion remains visible and balanced on both light and dark surfaces. Provider, category, recording, and task-handoff controls are keyboard operable, and scaled text switches work areas to a readable single-column layout. - **The desktop sidebar is navigation-first and keeps important task filters visible.** Saved task filters now appear directly under Tasks with live counts. The first five follow the order you set, More expands the complete list without a limit, and the manager handles reordering, renaming, and deleting. Recording, timer, and agent status share one compact expandable activity row, while healthy sync stays quiet until work needs attention. Saving a task filter now stays inside the filter flow, with explicit Update filter and Save as new choices instead of a separate popup. ## [0.9.1049] ### Added - **Lotti now guides first-time setup through one onboarding welcome.** The welcome walks through connecting an AI provider and creating a first task, replaces the older provider-selection prompt, and can always be replayed from Settings › Onboarding. ### Changed - **Survey questions now fit cleanly in their modal.** Bundled questionnaires keep every answer choice and the Next action visible on both phones and desktop instead of clipping the final controls in a fixed-height sheet. - **Inference profile model labels are readable again.** Capability names use normal caption spacing instead of widely tracked overline text, so labels such as Image Recognition no longer wrap character by character. - **Daily OS shutdown now stays usable on narrow phones.** Its actions stack responsively, and wrapped metric labels keep enough height at both standard and larger accessibility text sizes instead of overflowing their cards. - **Mobile navigation remains stable at fractional accessibility text sizes.** Bottom-bar labels now reserve the same rounded line height Flutter renders, avoiding tiny overflows at scales such as 1.3×. - **Time Analysis now stays usable in narrow windows.** Period shortcuts wrap below the navigator, headline cards stack in reading order, and chart controls move below their heading instead of overflowing or clipping. - **Completed Daily OS tasks keep their cover thumbnails.** Finished agenda rows now retain the same task artwork as open and in-progress work, making a busy day easier to scan without losing task identity after completion. - **Task agents now update only when you ask by default.** New agents keep watching their task without automatically spending inference tokens after every change. An outdated summary stays visible with a clear **Wake agent** action, while automatic two-minute updates remain available as an explicit top-level opt-in. The card now separates report reading from automation and model setup, and places **Read more** after the content it expands. ## [0.9.1048] ### Changed - **The remaining settings lists fade their dividers on hover.** Hovering a row in Categories, Labels, Habits, Measurables, Dashboards, or the Matrix sync maintenance list now fades the hairlines above and below it, so the hovered row is never bisected — matching Config Flags and Advanced Settings → Maintenance. The row layout stays fixed; only the divider colour changes. ### Added - **Desktop Lotti now has a keyboard-first command layer.** Open the command palette with Command/Ctrl+K, jump to the eight main destinations with fixed Command/Ctrl+1–8 shortcuts, move between desktop regions with F6, and use contextual save, search, refresh, and create commands in the primary work surfaces. macOS menus, command execution, and the fully localized shortcut reference in Settings now share one catalog, while keyboard users can resize split panes and confirm Daily OS commits without pointer-only gestures. ### Changed - **Logging a measurable now stays in one clear, accessible sheet.** Changing the observed date and time moves to a calendar page in the same capture flow instead of stacking another modal over the editor. Values and comments stay intact when navigating back, the complete time picker scrolls above the sticky action, and Save waits for persistence with visible progress and retry feedback. - **Task agents now use evidence-first processing for everyone.** The previously optional behavior is now standard and its Advanced Settings toggle has been removed. If you already had it enabled, nothing changes. - **Synced alerts are now available to everyone.** Alerts stay consistent across devices whenever Matrix sync is active, and the former Advanced Settings toggle has been removed. - **The Knowledge Graph is now available from every task.** The graph button is always present in normal task headers, while embedded task details suppress the recursive entry point. Graph controls, node and relation labels, relative ages, summaries, and error states are localized in all supported languages. - **Daily OS is now a schedule you can actively operate.** Long dictated plans stay fully visible and editable, AI activity appears from the first processing frame, and task-linked titles and categories update immediately. Planned blocks can be moved, resized, or edited in a design-system modal with the familiar Start & end controls, atomic saving, and Undo. - **Maintenance settings dividers fade on hover.** Hovering a row on the Advanced Settings → Maintenance page now fades the hairline dividers above and below it, so the hovered row is never bisected — matching the existing Config Flags behaviour. The row layout stays fixed; only the divider colour changes. - **The bottom navigation bar now gets out of the way on Settings detail pages (mobile).** Opening a terminal Settings destination — AI, Agents, any Sync page (backfill, stats, outbox, …), and Advanced pages like Config Flags, Logging, Maintenance, and About — slides the bottom bar off-screen so the page owns the full height, and it slides back when you tap back. Menu hubs (Advanced, Sync, Definitions) and the browse lists (Categories, Labels, Habits, …) keep the bar since you navigate onward from them. - **AI agents respond sooner when several are active.** Up to three different agent wake cycles now run at once by default instead of waiting in one global line, while each individual agent remains single-flight. AI Settings lets you tune the device-local limit from one through eight for local hardware or provider capacity. - **Task and project filters now share one coherent, accessible flow.** Status, category, label, and project choices stay inside the same smoothly resizing sheet or desktop dialog, with consistent full-width selection states, responsive actions, keyboard focus restoration, and clearer activity and sort controls. Linked-entry activity and sort pills now match the compact size of the task-header chips, with hover feedback confined to the pill itself; active activity pills retain their individual colored outlines. Task filters open immediately while project data refreshes in place, and active filters remain visible as compact chips on the task page. - **New tasks now stay in the filtered list where you create them.** A task created from a view with one category, project, or status selected inherits that selection, along with every selected label. Ambiguous multi-selections keep the existing defaults. ### Fixed - **Matrix sync now reliably catches up after a device was offline.** Reopening a device after another device sent a large burst now always walks forward from the last applied Matrix event instead of occasionally treating a cached timeline as complete and leaving newer entries missing. - **Command+S saves changed rich text again.** The save command now travels with the reusable editor, so it works in journal cards and task forms as well as on the standalone entry page without reaching for the Save button. - **Keyboard navigation now enters task and project details directly.** Right Arrow on a focused row opens the adjacent detail pane instead of resizing the divider, and keyboard row focus now uses the same full-width color and rounded clipping as pointer hover. - **Checklist cards now collapse cleanly after every item is complete.** The completion message stays on one line while the card closes instead of breaking into vertically stacked letters. - **The editor formatting toolbar renders correctly on macOS again.** Focusing an editor no longer replaces its formatting controls with a grey error strip. - **Bulk checklist suggestion acceptance no longer flickers.** Confirm all now keeps every proposal row mounted through its staggered exit, even when fast checklist check-offs resolve the underlying suggestions before later rows begin animating. The task page remains anchored throughout the batch. ## [0.9.1045] ### Changed - **Date and time pickers are clearer and steadier across the app.** Entry ranges now show weekdays, include Start/End **Now** actions, and move into a calendar page within the same sheet. Due dates and project target dates use the same weekday calendar, while tighter wheel controls and stable status spacing keep the layout compact without jumping in light or dark mode. - **Pickers now feel like one coherent part of Lotti.** Task status, priority, labels, category and project, project status, AI profiles and models, and Gemini thinking mode now share the same full-width selection rows, selected state, typography, hover/focus treatment, and adaptive phone-sheet or desktop dialog presentation. Task-agent AI setup keeps profile, provider, and model choices inside one navigable sheet, while pausing AI uses an inline confirmation instead of opening another modal. Its overview now groups the active profile, effective model, and automation into a compact settings hierarchy, and keeps the loaded AI catalog between pages so navigation no longer flashes an empty state. Page changes now use a calm staggered fade while the sheet eases smoothly to its new height, and a tapped thinking model shows its selected state immediately while the change is saved. - **Mistral Voxtral now supports contextual chat transcription.** Instruction-following Voxtral Mini and Small models use the same temporary MP3 path as Melious, so long recordings stay compact while task instructions and speech-dictionary terms are applied during recognition. The archived M4A remains unchanged and temporary files are removed after every request. ### Fixed - **Accepting checklist suggestions no longer makes the task page jitter.** Single and bulk accepts now keep the suggestion area anchored before each checklist layout change is painted, even when several to-dos arrive across successive refreshes. ## [0.9.1044] ### Changed - **Daily OS now lets you choose its AI route explicitly.** A discoverable Daily OS settings page owns the default inference profile and preferred name, while the planner instance can override its full profile or only its thinking model. The app states that assembled planning context is sent to the selected provider, identifies local versus remote endpoints, and guides you to setup before the first check-in when no usable profile is configured. - **Your Daily OS name now follows you across devices.** The preferred name you set for Daily OS syncs to your other devices, with the most recent edit winning, so each device greets you the same way. ### Fixed - **Long Voxtral recordings transcribe without oversized WAV uploads.** Melious Voxtral requests now encode a temporary MP3 from the compact M4A recording, keeping the archived audio unchanged while avoiding the request-size failures that affected recordings longer than roughly one or two minutes. Temporary decoder and MP3 files are removed after both successful and failed requests. - **AI coding prompts now show up under their task.** A generated coding prompt is linked to both its parent task and the audio or text note it came from, and appears in each of their linked-entries lists (under the Code activity filter). Previously the prompt was created but stayed hidden from the task timeline. ## [0.9.1043] ### Changed - **Daily OS is now available to everyone.** The Daily OS day-planning surface no longer sits behind a Settings → Advanced → Flags toggle — its calendar tab appears in the main navigation for all users. ### Fixed - **Sync settings pages now slide back consistently.** Returning from a Sync sub-page (such as Backfill sync) to Sync settings uses the same clean pop animation as every other settings screen, instead of the jarring page-swap transition it played before. ## [0.9.1042] ### Added - **Pause and resume audio recordings mid-capture.** The recording sheet now has a pause control next to stop and cancel, so you can stop talking to gather your thoughts and pick the recording back up without splitting it into separate entries. While paused, the elapsed timer and level meter freeze; resuming continues the same recording. (Available for standard recordings; live-transcription recordings are unaffected.) ## [0.9.1041] ### Changed - **Task details stay steady while AI results and accepted suggestions land.** Completed audio transcripts and image analyses now expand smoothly instead of snapping linked entries taller, and content already above the viewport no longer yanks the passage you are reading. Checklist items created from an accepted AI suggestion also reveal progressively while the suggestion card remains anchored under your pointer. ### Fixed - **Voice input now reports transcription failures instead of silently disappearing.** Chat, saved audio entries, linked tasks, and agent-evolution input show provider diagnostics, including empty-response failures. Melious Voxtral now converts a temporary copy of the compact M4A master to WAV for a buffered contextual chat request on every supported platform, then deletes its scratch files. Conversion uses native system media APIs without bundling FFmpeg, and failures surface their decoder detail instead of dropping to a less contextual request. Requests fail with provider detail and a request ID after 60 seconds instead of waiting indefinitely. Voxtral is preferred over Whisper when both Melious models are configured. ## [0.9.1040] ### Added - **Task agents now show and own their AI setup directly in the summary.** The header names the thinking model, its publisher, and the serving provider; tapping it opens one persistent setup surface for choosing a profile or model, returning to the profile default, copying the category default, or pausing AI entirely. Reports keep immutable model/provider attribution after later setup changes. Automatic task-related updates can also be switched off without removing the manual Run now action or the existing report. ### Fixed - **AI profile and model choices now use one consistent picker flow.** Task-agent profile and model changes finish saving before the setup sheet closes, and their confirmation stays inside the task column instead of spanning the app. Category defaults, agent creation, agent templates, inference-profile slots, and task-agent setup now share the same design-system selection language. - **Backfill Sync now shows its progress as it happens.** The Missing counters in the status row and statistics ledger update as sequence-log rows resolve, keeping pace with the inbound queue countdown during large backfills instead of waiting for the slower diagnostics refresh. - **The energy-orb recording visualization no longer overwhelms Linux rendering.** The orb now compiles only its production shader and reuses the loaded fragment shader as live audio levels change, avoiding the prolonged Settings stall and recording crash seen on affected Linux graphics stacks. Finishing a recording also dismisses its sheet exactly once before opening the saved entry, avoiding a nested-navigator teardown failure. ## [0.9.1039] ### Added - **Task agents now use evidence-first inference by default with efficient models.** The mode remains configurable in Settings → Advanced → Flags and uses an evaluated low-variance, temperature-zero path. The default Melious profile now uses Qwen 3.5 122B A10B directly for task reasoning and keeps Mistral Small 4 119B for image recognition. Existing untouched Melious profiles move to Qwen automatically, while customized model choices remain unchanged. Qwen reports that match known harmful regression patterns receive a compact, bounded Qwen repair; clean reports remain single-pass. Selecting Mistral instead keeps Mistral responsible for task and checklist changes and adds an isolated Qwen report pass. Evolved and manually customized report instructions remain authoritative, each model call is tracked separately, and concrete multi-step plans become checklist actions without requiring a separate explicit checklist request. ### Fixed - **Melious task agents now continue reliably after calling tools.** Follow-up requests now preserve the OpenAI-compatible assistant message shape expected by strict providers, so checklist updates and the final report are no longer interrupted after an otherwise successful tool call. - **Task-agent reports stay current after task changes.** When an agent successfully changes task metadata or checklist state but omits its report call, Lotti requests a fresh report instead of continuing to display the previous state. True no-op wakes still reuse the existing report. Agent internals also record whether the Qwen report pass was accepted, rejected, failed, not needed, or ineligible. - **Failed task-agent inference is no longer recorded as a successful empty wake.** Provider and connection failures now reach the workflow failure path so retries and diagnostics reflect what actually happened. ## [0.9.1037] ### Added - **Daily OS gives better feedback while turning a spoken check-in into a plan.** Drafting now shows a staged progress trail, failures offer Retry and Back to decisions, the Day footer explains why blocks made it in, each block can show its own reason, and quick review actions let you accept the plan or ask for less, a lighter move, or more buffer without re-dictating from scratch. At larger text sizes, those refinements collapse into one Adjust menu so the footer stays reachable. - **Share AI Impact without leaking your category names.** The AI Impact breakdown now has a **Both / By category / By model** picker. **By model** hides every category name, so you can screenshot and share your model mix and its cost without exposing what your categories are called. - **Interactive AI Impact charts (favorite models over time).** The AI Impact dashboard now visualizes the breakdown, not just tables. Separate **by category** and **by model** charts are shown together, each with a **Share** mode for how your mix shifts week over week (the model chart is the Cursor- style favorite-models view). A **Requests** metric and per-model unit economics (calls and cost per million tokens) join the KPIs, whose tiles are now the metric selector — tap a tile to drive the charts, and the selected tile shows how it moved versus the previous period (e.g. "▲ 24% vs May"), with rising cost/energy/CO₂e flagged in a caution colour. The charts are interactive: **tap a series** (in the legend or a breakdown-table row) to isolate its trend, and **tap a bar** to scope the "Recent calls" list to that period. A note makes clear that cost, energy and CO₂e are measured for cloud models only. - **Better AI Impact accounting for transcription.** The AI Impact dashboard now has a **Running total** chart mode for cumulative cost, energy, CO₂e and token usage over the selected period. Voxtral transcription records token usage when the server returns it in final accounting chunks, and transcription calls whose provider reports no token, cost or energy metrics are labelled instead of leaving a blank ledger row. - **AI Impact by model and serving location.** AI Impact now adds a model breakdown for the selected metric, using provider-native model IDs when available. When the provider reports data-center metadata, it also groups environmental impact by country/data center and shows the renewable-energy share for each location. - **Pick your recording look in Settings.** A new Settings → Recording Style page lets you choose the VU meter or energy-orb visualization for the mic — with the same live preview and "Try with your voice" toggle from onboarding — and the real recording screen now honors your choice instead of always showing the VU meter. ### Changed - **The next-gen agentic DailyOS is now the only DailyOS.** The "Use next-gen agentic DailyOS" toggle under Settings → Advanced → Config Flags is gone: the calendar tab always opens the voice-first, agent-led capture and reconcile flow, and the previous DailyOS surface (including its Set Time Blocks screen) has been removed. - **AI profile pickers only show profiles you can actually use.** Default inference profiles are no longer seeded for providers you haven't set up: each bundled profile (Gemini, OpenAI, Mistral, Melious.ai, Chinese AI, Anthropic, Ollama, oMLX) now appears only once a working provider of its type exists — connect Melious.ai and you get exactly the Melious profile, nothing more. Untouched leftover profiles for providers that were never set up (or have since been deleted) are cleaned up on the next launch; any profile you renamed, described, or rewired to another provider is kept. ### Fixed - **Dashboard chart editing now saves when it should.** Adding measurable charts in Settings → Dashboards immediately enables Save, measurable chart aggregation can be chosen before adding, aggregation and reorder no-ops no longer make the form look changed, chart rows expose clearer edit/remove and reorder affordances, and the chart sections and pickers now use the shared design-system layout. - **Daily OS dictated plans keep task agents and titles in sync.** Tasks created from a spoken Daily OS plan now get the category's default task agent, and task-backed planned blocks on the Day timeline update when you rename the task. - **The desktop sidebar is calmer when Tasks has lots of saved filters.** Saved task filters now sit in their own capped Tasks section with a More row, tapping More reveals every saved filter, active filters stay visible, recording/timer/agent cards carry clearer status labels, agent rows can open their linked task directly, and the old tiny Outbox/Inbox footer is now one compact Sync strip. Daily OS and Insights sub-items now use the same grouped sidebar surface, so Time Analysis and AI Impact read as intentional secondary destinations instead of loose rows. - **Keyboard and screen-reader flow is clearer across shared desktop controls.** Saved task-filter rows can now be focused, opened with Enter or Space, renamed with F2, cancelled with Escape, and deleted with a two-step Delete/Backspace flow. Screen readers announce the row as a selected button with rename/delete actions, hidden delete controls stay out of the focus tree, glass overlay buttons now carry proper labels/tooltips, modal Back/Close controls are named, and live sidebar cards announce as actionable buttons with their linked title and elapsed time. - **Habit and agent-heavy screens spend less time waiting on the database.** Habit heatmaps now load only the latest completion per habit/day, and agent summary/feedback paths batch payload lookups instead of firing many individual reads. - **Entry timestamps no longer get cut off on phones.** In a task's linked entries, the header's date and time used to be truncated (e.g. `2026-07-08 14…`) when the action buttons squeezed the row on a narrow phone. The timestamp now stacks the date over the time on two lines when space is tight, so the whole date and time stay readable. - **Choosing an AI profile or model now works on phones.** On phones, tapping a choice in an AI picker — a category's default inference profile, a profile slot's model, or an agent template or soul — closed the whole settings page instead of the picker, so your selection was never saved. The picker now closes on its own and keeps your choice. - **Melious Voxtral calls now show real AI Impact usage.** Melious-hosted Voxtral audio models now use the chat audio endpoint, which reports token usage, cost, and environmental impact, instead of the transcription endpoint that only returns text. Category speech-dictionary terms are included in the chat prompt so Voxtral still gets spelling context. - **The image viewer's save button now works everywhere.** On iPhone and Android it saves the photo to your photo library; on macOS, Windows and Linux it opens a native save dialog so you choose where the image goes. Previously it always tried to write into `Downloads/Lotti`, which failed with an error on iPhone and on the sandboxed macOS build. - The debug-only Onboarding animation gallery (Settings → Advanced → Maintenance) no longer leaves the floating bottom navigation bar stacked on top of it on phones. It now escapes onto the root navigator the same way other pushed debug/editor surfaces already do. ## [0.9.1032] ### Added - **A calmer image viewer.** Tapping a journal photo now opens it over a dimmed version of the app instead of a stark full-screen shell. The viewer keeps a visible margin around the photo, has persistent top-right Download and Close buttons, closes with Escape, and adds a bottom zoom control while preserving pinch-to-zoom and panning. Download saves the original image into `Downloads/Lotti` with a unique filename. ## [0.9.1031] ### Added - **See what your AI actually costs — in money, energy, and CO₂e.** Lotti now records every AI call it makes (agents turn by turn, transcription, image analysis, cover art, text generation) with its cost in credits, energy use, CO₂ footprint, water use, tokens, and the data center it ran in — measured per call by Melious for Melious-hosted models. A new **AI Impact** dashboard (next to Time Analysis in the sidebar, and under Settings → AI → Usage & Impact) shows where it goes: switch between Cost, Energy, CO₂e and Tokens, browse by week, month or year, and see stacked per-category charts, period totals, a ranked category breakdown, and a ledger of your most recent individual calls (time, model, kind, tokens, cost, energy). Tasks that used AI carry a quiet header chip with their lifetime footprint — for example "€0.42 · 12 Wh · 3.4 g" — with the full breakdown a hover away. Consumption records sync across your devices, so the numbers cover everything, not just one machine. - **Cancel a recording.** The audio recording sheet now has a discard (✕) button next to Stop, in both standard and realtime modes. Tapping it throws the recording away entirely after confirmation — no audio entry is saved, no transcription runs and no task summary is generated — and returns you to the page exactly as it was before you started. Stop still behaves as before: it keeps the recording and kicks off transcription and the linked task's agent. - **Saved task filters on your phone, synced across devices.** The named filter presets you build on the Tasks page now come to mobile. A slim rail under the Tasks header shows the filter you're on and a couple of recent ones with live task counts and one-tap switching; tapping **Saved** opens a "Saved filters" sheet — the full list where you pick one to switch, hit **Edit** to rename or delete, or **Save current filter as…** to keep a view you've tuned. When you've hand-tuned an unsaved filter, a **Custom** pill and an inline **+ Save** appear right on the rail. Saved filters now sync between your devices too: create or delete one anywhere and it shows up (or disappears) on the rest, and a new **Settings → Sync → Sync Entities → "Saved task filters"** step pushes across the presets you already had. Each device works out its own counts, so the numbers always match what that device actually holds. - **Customizable completion celebrations.** Settings → Advanced → Animations now lets you tune each celebration style — particle count, size, spread, and per-style physics like gravity, twinkle, sway, spin, swell — in a full-screen playground. Controls are grouped into Shape / Motion / Look, each with a plain-language description, and you can drag a slider or tap the value to type an exact number. A live preview plays the effect inside a sample checklist (so you see how it reads in context, not over a lone dot), and Reset-to-default can be undone. Changes persist and apply wherever that style plays. Each content type can also be set to **Random** (a different style on every completion) or **Combine two** (two styles layered, re-rolled each time) for more novelty. - A guided **first-time onboarding** (behind the "New onboarding (FTUE)" flag in Settings → Advanced → Flags). With the flag on, a fresh install replaces the bare provider picker with a short, skippable, cinematic flow: a welcome that frames AI setup as "connect your brain"; a provider step (Melious.ai first, then Mistral, Gemini and Qwen, with OpenAI and Ollama behind "More options") that pastes a key and creates the provider, its known models and a ready-to-use inference profile in place; a quiet "connected" beat; a step to pick the life areas (Work, Fitness, Family, Friends, or your own) the AI should power — teaching that the provider is chosen per category; and a playful step to pick how recording looks (a modern energy orb or an analogue VU meter, previewed live — try it with your voice). It ends in the real payoff, without ever leaving the onboarding dialogue: a "Create your first task" step in the same panel where you speak a thought — recorded with the visual you just picked — or tap a starter suggestion (like "Plan my week") if you're not ready to talk. Your words become a structured task with a checklist, revealed right there in the panel as a glowing card ("Your first task is ready"); tapping it lands you straight on that task — already in progress, in the area you picked, with your original voice recording attached to it, and with Laura, the default task agent, already assigned and immediately at work — her first pass over your words lands as a summary with suggestions for you to confirm, just like after recording a voice note inside a task. Every area you pick during onboarding gets Laura as its default agent, so later tasks in those areas get an agent too. Areas you already have (like an existing "Work" category) are reused instead of duplicated — rewired to the new provider, but keeping any agent template you had chosen yourself. A "Rather type?" path, a one-tap retry, and a soft-landing title-only task when structuring fails keep it robust, and the whole flow honors reduce motion. - A first-class **Events** experience (behind the "Enable Events" flag in Settings → Advanced → Flags). Events — birthdays, trips, milestones — now get their own destination (under *More* on mobile, the sidebar on desktop) with a photo-led overview that groups your moments by time and filters by category, plus a redesigned detail page that doubles as a full inline editor: a photographic header where you rename the event, set its category, status and rating, and add or change a cover; an at-a-glance summary; a photo gallery you can tap into full-screen and swipe through; a timeline of the linked photos, notes, voice notes and time recordings (each opening its source entry; a time recording reads as a span with its elapsed duration); and the prep/follow-up tasks tied to the event. Add photos, notes, voice memos and tasks straight from the page — even a brand-new event shows what to do next instead of an empty screen. - An optional **AI recap** for events. Point a category at an event-agent template (Settings → Categories → AI defaults) and new events in that category get a quiet writing companion: once the event has a photo or a note, it weaves the linked photos, notes and voice memos into a short, warm recap shown on the event's detail page, with a refresh to rewrite it. It can also propose the occasional concrete follow-up — a thank-you to send, an album to share — which you accept or dismiss with a swipe; accepting turns it into a task linked to the event. It never touches your star rating or cover photo — those stay entirely yours. - AI settings now include Melious.ai as a dedicated EU-hosted inference provider. The provider setup can fetch Melious' live model catalog, including chat, reasoning, vision, audio, image, embedding and rerank models, and add any returned model as a saved Lotti model row instead of relying on a hard-coded static list. Large live catalogs are searchable in the Available Models section instead of expanding into one long page. New Melious setups also seed a ready-to-use default profile with Mistral Small 4 119B Instruct for thinking and image recognition, GLM 5.2 (a 1M-token-context reasoning model) as the high-end thinking option, Flux 2 Klein 9B for image generation, and Voxtral Small 24B for transcription — a speech-recognition model that takes your category's speech dictionary into account, so names and domain vocabulary come out right where Whisper guessed. Whisper Large v3 and Turbo are still created alongside for easy switching, and existing untouched Melious profiles move to the new GLM 5.2 and Voxtral defaults automatically. - Deleting an AI provider and reconnecting it later no longer leaves its profile silently broken. The seeded profile used to keep pointing at the deleted provider's model entries, so everything that resolves through it — structuring your first task, transcription, and agents like Laura — failed with "no inference provider configured" even after you added the provider back. Broken profile entries now heal automatically: on every app start and immediately after connecting a provider (in onboarding or settings), so your first capture works seconds after the key step. Model choices you made yourself are never touched — only pointers that lead nowhere are repaired. - oMLX can now be used for local Whisper Large v3 transcription on Apple Silicon. The oMLX model catalog includes Whisper Large v3, Whisper Large v3 MLX, and Whisper Large v3 Turbo as audio-to-text models, and selecting one routes voice-note transcription through the local OpenAI-compatible `/audio/transcriptions` endpoint instead of chat completions. Untouched local oMLX default profiles now use Whisper Large v3 Turbo for transcription. - oMLX provider settings now fetch the local `/models` endpoint in the Available Models section, so any model exposed by the running oMLX server can be installed directly from the provider edit page. - Mistral provider settings now fetch the live `/v1/models` catalog in the Available Models section, mapping each model's capabilities (vision, OCR, audio transcription, reasoning, function calling) so you can search and install any Mistral model — not just the curated defaults — directly from the provider page. - Gemini and OpenAI provider settings now fetch their live model catalogs in the Available Models section too, just like Melious, Mistral, and oMLX. Gemini reads Google's native `/v1beta/models` listing (display names, descriptions, token limits and the thinking flag); OpenAI reads `/v1/models` and derives each model's capabilities from its id, dropping embedding, moderation and text-to-speech rows. When a new Gemini or OpenAI model ships, you can search for it and install it straight from the provider page instead of waiting for a curated update. - **Mistral OCR support.** Running "Analyze Image" with a Mistral OCR model (e.g. `mistral-ocr-latest`) now extracts the document's actual text. OCR models use Mistral's dedicated `/v1/ocr` endpoint instead of chat completions (which rejected them), so a scanned letter or form comes back as real Markdown text you can read and ask follow-up questions about — rather than a vision model's loose description. ### Changed - Provider detail pages now list your installed models above the searchable model catalog, so you see what you already have before scrolling on to add more (Melious, Mistral, and oMLX). - On platforms with HEIC/HEIF conversion support, image import now accepts high-efficiency inputs from drag-and-drop, file picker, and clipboard paste. Lotti preserves existing JPEG and PNG inputs as-is, converts ordinary HEIC/HEIF files to JPEG, and converts HEIC/HEIF files with alpha to PNG so transparent screenshots stay transparent. - **The FTUE setup animation feels more organic.** The welcome screen now grows an entangled, looping neural organism with dim tendrils, smaller offshoots and travelling activation tips instead of separate straight-line constellation clusters. Later setup steps keep a softer living backdrop, a lighter cloudy modal surface and calmer provider cards so the flow feels connected without fighting the form content. - **Choosing a model for a one-off AI run is calmer.** When you run an AI action from the sparkle menu (transcribe a voice note, analyze a photo, generate a prompt or cover art) and you have more than one capable model, the picker no longer drops you into one long flat list. It now groups your models **by provider** — pick a provider (Anthropic, OpenAI, Gemini, Mistral, a local Ollama…), then a model within it — with a one-tap shortcut to your current default pinned on top. If you only have one capable model it runs straight away, and if all your models live under a single provider the picker skips straight to them. **Cover-art generation now lets you pick which image model to use, too** — previously it always used your profile's image model. None of this changes your saved profile; the choice applies to that one run. - **Flux cover art gets its own prompt path.** The AI popup now includes a separate *Generate Cover Art (Flux)* action for task-linked notes and voice notes. It sends Flux models a short visual scene and mood brief instead of the full task JSON and related-task context, while keeping the same 16:9, center-safe thumbnail composition guidance as the regular cover-art prompt. Melious-hosted Flux 2 Klein 9B gets a clean image prompt for task covers, requests an explicit 16:9 `1792` x `1008` canvas through Melious' FLUX dimensions, and new and untouched Melious profiles use it as their default image-generation model. - **AI settings search now matches the rest of the app.** The search field in Settings → AI uses the same design-system search component as the journal, habits, task-linking and language pickers — one consistent look instead of a one-off purple-tinted field — and its placeholder now reads *Search providers, models, profiles…* instead of the internal *Search AI configurations…*. - **The AI settings "add" button and card menus are clearer and easier to hit.** The floating add button shows one unmistakable **+** on every tab (instead of three different icons), the **⋯** menu on each provider / model / profile card is a full-size touch target, and the add button no longer hides the last item in the list on phones. - **Settings controls now speak the same visual language as the rest of the app.** Save and Cancel buttons across the settings pages — AI providers, models and profiles, sync and labels — use the design-system button: quiet and greyed when there's nothing to save, waking to a teal accent the moment you change something (the same pattern as the note editor). The toggles, checkboxes and the light / auto / dark theme switch use the design-system controls too, with teal accents and consistent shapes in place of the old purple Material widgets. The category chips on a label now carry their category's colour as a soft tint (instead of a heavy solid fill), and small status tags — a profile's "Desktop only", a sync node's capabilities — use the design-system badge. The AI model filters (provider, capability and reasoning) on the Models tab are design-system chips now, each provider keeping its colour on a small avatar dot. The Sync conflicts and outbox lists now sit on the same dark background as the rest of Settings instead of a near-black sheet (with their cards reading as elevated), and the Sync Outbox row shows its pending count as a clean badge rather than a postbox glyph with the number squeezed on top. - **Accepting or dismissing an AI suggestion is now smooth and calm instead of jumpy.** The row you tap stays put and confirms in place — a clear filled badge with the plain word *Confirmed* or *Dismissed* and a light haptic — then gently collapses while the rest of the list slides up to close the gap; the pending count ticks down with the action. There's no longer a success popup at all (for single accepts or "Confirm all") — the motion makes success abundantly clear. "Confirm all" sweeps the list clean in one staggered motion with a single tap-feel. The swipe hint no longer nags: it plays once, as a single gentle nudge, instead of a repeating two-way wiggle. While a row is leaving, the rows sliding up briefly ignore taps so a fast second tap can't hit the wrong suggestion. Screen readers hear the verdict and how many are left; with *reduce motion* on, rows update instantly with no sliding. - Checking off a checklist item now plays the same little spark burst no matter how it happens — a direct tap, an accepted AI "check off" suggestion, or a sync from another device — so completing an item always feels rewarding. - **Completion celebrations are now yours to shape.** A new master switch in Settings → Advanced → Animations turns every completion flourish on or off at once, and you can give each kind of moment its own style — choose from Sparks, Fireworks, Confetti, Embers and Bubbles for finishing a task, keeping a habit, and ticking off a checklist item (they start out as sparks, confetti and bubbles). Pick a surface to see the style it's set to, and the picker plays a preview as you choose; a "Try it" row fires each one on a dummy Done button, checklist item and habit before you commit. And the completion buzz now has its own switch, independent of the animation — keep the haptic with the flash off, or the flash with the buzz off. - Dashboard charts with no data in the selected range now collapse to a compact one-line notice under their title instead of holding open a full-height empty card. On a dashboard with several charts, the ones that actually have data get the room, and you no longer scroll past tall blank panels to reach them. - **Editing a note now has a calmer, clearer toolbar.** Saving moved out of the card footer and into the editor toolbar: a pinned Save button that holds one spot the whole time you edit — quiet and greyed when there's nothing to save, and waking to a teal accent the moment you change something (the disk icon is always there, so the change reads as more than just colour). When you have unsaved edits, a discard control appears next to Save to throw them away and drop back to the saved text. The toolbar also adapts to width: on a roomy desktop window every formatting control is one click away, while on narrow screens it shows the essentials plus a "…" button that opens the rest. - **New AI suggestions now ease into place instead of snapping.** When a fresh proposal lands, its row grows open and the content below it glides down rather than jumping — and if the card has scrolled off the top of the screen, nothing on screen moves at all. The event recap and proposal cards and the daily-planning "learning" cards reveal the same calm way. - **Charts no longer replay their grow-in animation on every refresh.** The dashboard bar chart and the habits completion-rate chart used to re-animate each time their data refreshed or you changed the time range; now they update in place. - **Reduce motion quiets more of the app.** With *reduce motion* on, the looping pulse on the pending-review dot, the rate-session button and the voice-capture orb hold steady instead of breathing, and the full-screen "locking in" takeover resolves at once. - Scrolling the time-history strip at the top of the Day view is smoother on long histories — the background chart no longer rebuilds on every frame as you drag. - **The template and soul evolution chats no longer jerk to the bottom while you're reading.** If you've scrolled up to look back at earlier messages, a new reply (or the typing indicator) no longer yanks you down — the view only follows along when you're already at the bottom, and eases there instead of snapping. - **Dragging a planned block on the Day timeline now settles smoothly.** When you let go, the block eases into its snapped 15-minute slot instead of teleporting to it. - Checklist cards in the journal feed no longer grow a step taller a moment after they appear — their progress chip and bar hold their space from the first frame, so the list doesn't shuffle under you as you scroll. ### Fixed - Linked tasks now show the actual task status glyph. In-progress linked tasks use the same play icon as the task status pill and status picker instead of falling back to an open circle, and the link-task picker uses that same shared status icon set. - Task cover art no longer flickers or briefly disappears while resizing the desktop panes. Dragging any pane divider — including the sidebar, away from the detail page — used to re-decode the cover image on every frame of the drag; the detail page now reuses the already-decoded image while the width changes and keeps the current frame on screen when a genuinely larger decode is needed. - Entry header action buttons (the favorite star and the `…` overflow menu) are no longer clipped at the top-right on narrow phones. When the row ran out of room the timestamp now shortens instead of pushing the buttons off the edge. - Coding prompts generated from a task-linked audio note now attach to the parent task even when the note's task relationship has to be resolved from the link graph, so follow-up prompts can see the earlier generated prompt as task context. - Photos dragged into Lotti now keep the moment they were taken. They were being dated to the time of the drop; the importer now reads the original capture time from the image's EXIF metadata (and its GPS location, when present), falling back to the file's modified time only when the photo carries no timestamp. That fallback used to read as the drop time too, so dragging a batch of screenshots from a meeting stamped them all with the same time and out of order — dragged files are now read straight from their original location on disk instead of a copy, so the real modified time (and the original order) comes through. - The celebration **Style** settings got a small polish pass. Each style's preview is now a single crisp play button instead of a blurry triangle and dot layered on top of each other, and the Tasks / Habits / Checklist items selector above it uses the same aligned segmented control as the rest of the app, so its rounded corners line up with the selected segment. - An inference profile whose model's provider had been deleted no longer shows a garbled raw model id in its slots. The slot now reads "Model unavailable — its provider may have been removed" and keeps its clear (×) button so you can reset the dangling reference. - On phones, the back button on a settings sub-page (e.g. Advanced → Maintenance) could fail to return to the previous screen. It now follows the visible navigation stack — the same as the system back gesture. ## [0.9.1030] ### Changed - The desktop sidebar's live-status area — the running timer, an active voice recording, and background agents — was redesigned to feel calmer and more cohesive while still reading at a glance. The old jumble of red, bordered, and monospace boxes is gone. A running timer is now a teal card and an active recording a red card, each with a coloured accent rail, the task name on up to two lines, and a big, easy-to-read elapsed time in its accent colour; the recording shows a gentle pulsing dot so you can tell it's live (it respects the system "reduce motion" setting). Background "Agents" sit just below in a quieter card. Hovering a shortened task name shows it in full, and at the very bottom the sync readout is now labelled "Outbox / Inbox" (instead of "tx / rx") and stays quiet until there's something to sync. - The **entry cards on a task's timeline** got a whitespace-and-typography pass so the different entry types (text, audio, image, measurement, health, workout, habit, rating) read as one calm, legible family. The timestamp is now a quiet but readable caption (it used to be a faint, monospace-looking stamp), and every "label: value" line shares one treatment — a quiet label with the number itself as the bold, high-contrast value — left-aligned to a single content gutter. Measurements no longer print their note twice, "55 %" reads "55%", health/measurement values stop floating centered, the rating card now shows a number next to each bar (e.g. 80%), and faint header/footer controls (the timestamp, duration, overflow ⋯, expand chevron, audio timecodes, the AI button) were lifted to a legible contrast. Cards also hug their content with one symmetric padding instead of reserving an empty band at the bottom. - The task details **header** was reworked for a calmer, more responsive layout. The status control no longer floats in empty space — it leads a two-lane metadata block (status · priority · due · time estimate on the first lane, your labels on the second), with status and the due date carrying more weight than priority and tracked time. Every chip shares one height and style, the breadcrumb sits as a quiet eyebrow above the title, and the title binds tightly to its metadata. The title is now click-to-edit (no separate pencil button), and on narrow phones the due date and time estimate stay paired on one row. Chips read more clearly too: priority is spelled out (Urgent / High / Medium / Low) instead of the cryptic "P2", the time-estimate chip reads "tracked of estimate" in plain duration units (e.g. "0m of 1h" rather than a clock-like "00:00 / 01:00") with a tooltip, neutral metadata chips carry a defined outline so their edges stay legible, and a long label list collapses behind a tappable "+N" chip. The layout holds up across long wrapping titles, many labels, empty new tasks, narrow widths, and tasks with cover art. - Two spots on the task page are easier to hit precisely. The AI suggestions' reject control is now clearly separated from accept, so a mis-tap can't trigger the wrong one. Each Todo checkbox gets a 44px tap target that sits in a subtle outlined well, so the whole tappable area is visible before you touch it (and lights up on hover/press) — no more aiming at the tiny square. The drag-grip dots are quieter so they don't compete with the task text, and the empty checkbox is drawn with a higher-contrast outline. - The AI summary card reads as a slightly more vivid "assistant" panel — a touch more accent in its corner wash and glow so it no longer looks muted — with evened-up spacing: a smaller gap above (matching the other section gaps) and real breathing room below before the bottom action bar. - The status / priority / label picker modals are tidier and more consistent: each row highlights with a rounded, inset shape on hover or selection instead of a sharp edge-to-edge band, and the titles no longer mix styles (the stray trailing colon on the status picker is gone). - Sync conflicts got a ground-up redesign. When the same entry was edited on two devices, you now see a full field-by-field comparison — title, body, category, dates, flags and more — with the exact words that changed highlighted, so you can tell what a choice would keep or discard instead of guessing. You can keep either version or **combine** them field by field, and a "recommended" hint points at the option that loses nothing. A "deleted on one device, edited on another" clash is shown as a clear, safe choice rather than a confusing diff. And new conflicts now raise a notification, so you find out right away instead of having to dig through settings. - The sync outbox page (Settings → Sync → Outbox) was redesigned to be clear instead of cryptic. It now opens with a plain-language summary — "Everything's synced", "Sending 3…", or "2 couldn't send" — and items that failed to send finally explain themselves and reassure you they're still saved on this device. You can retry a single item or retry everything at once, and each item's technical details (payload size, retry count) stay tucked away until you tap its card, so the everyday view stays calm. Removing a queued item now clearly warns that the change won't reach your other devices. ### Fixed - Accepting an AI-suggested label now works even when the task already has labels. Previously, once a task carried three labels — whether you added them yourself or accepted earlier suggestions — any further suggested label was silently dropped on accept. Suggestions are now applied regardless of how many labels the task already has, while a label you already have is still skipped as a duplicate. - Accepting an AI suggestion that adds a to-do no longer makes the page jump: the task detail view holds the AI proposals in place when a confirmed change grows the checklist above them, so the spot you were reading stays put. - Accepting an AI suggestion no longer makes the checklist flicker: a reloading item row keeps showing its current state instead of blanking for a frame (stale-while-revalidate), and checklist cards are keyed by identity rather than list position, so inserting or reordering one no longer drops the others' state and flashes them. - Adding a checklist item no longer flashes a missing divider: the line that separates the new item from the one above it is now part of the new row itself, so the item and its divider appear together instead of the divider arriving a frame later. - Confirming an AI proposal that checks off a checklist item no longer slides the AI card up from under your finger. The checked item's row collapses a beat after the tap, and the card now stays pinned across that delayed collapse, so you can keep tapping in the same spot to accept the next proposal. The pin steps aside the instant you scroll. ## [0.9.1029] ### Added - AI settings now include oMLX as a first-class local OpenAI-compatible provider. The Local Power profile now defaults to Qwen 3.6 35B-A3B 4-bit via oMLX for both reasoning and image recognition. Sync-node capability detection also advertises reachable local oMLX instances for profile pinning. - Added a separate Local Gemma 4 (oMLX) profile backed by `gemma-4-26B-A4B-it-QAT-MLX-4bit` for both reasoning and image recognition, keeping the Qwen-backed Local Power profile available as its own option. ### Changed - Task and project agents now send leaner, more cache-friendly prompts to local models: open proposal details are no longer repeated in the same wake, linked task report timestamps are omitted, and trigger-token lists are rendered in a stable order. - The experimental task knowledge graph now keeps expanding as you walk through project tasks. When you walk from a project to one of its tasks, Lotti loads that task's own linked neighborhood, adds the new nodes and links to the graph already on screen, and recomputes the merged layout so the world stays explorable instead of turning sibling tasks into dead ends. ## [0.9.1028] ### Changed - The Task details page got a high-end, celebratory rework. The task title is now the clear focal point, with the breadcrumb and unset fields quietly de-emphasised but still legible. The AI "Proposed changes" read as clean, colour-coded cards — on a phone each shows its change type above full-width text with visible accept / reject buttons (and you can still swipe right to confirm, left to dismiss), and the AI card itself now feels like a crafted assistant surface with a glowing sparkle badge and a clear "Confirm all" button. Your actual work — the checklists — now sits above the AI suggestions. Completing things is celebrated: checking a checklist item gives a gentle pop and a left-to-right strike-through, and finishing a whole checklist or marking the task Done blooms a soft glow with a ring of sparks around the progress ring / status pill (plus a haptic), tuned so it never covers the text it's celebrating. Sections fade in gently as the page opens. Everything honours the system "reduce motion" setting — the celebrations become a quiet static acknowledgement with no flying particles — and on wide screens the content sits in a comfortable centred reading column. - You can now switch the completion celebrations off. Settings → Advanced → Animations has three independent switches — habit completion, checklist-item completion, and task completion — so you can turn off the glow, sparks, and pop for any of them you'd rather not see. The brief haptic on completion is unaffected, and completion itself works exactly the same; only the animation is skipped. - The experimental task knowledge graph now feels more physical: tapping or walking to a node turns the focused 1–2-hop neighborhood into a short-lived local force simulation, with edge springs, local separation, damping, and a weak anchor back to the static layout. Walked-to nodes now get a stronger, lower-damping wobble so they keep calling attention after the camera lands. Desktop trackpads can now zoom the graph with a two-finger scroll gesture. Links stretch and settle without recomputing the graph, and the motion stops once settled and respects the system "reduce motion" setting. ## [0.9.1027] ### Changed - The logbook (journal list) was redesigned to match the rest of the app. Every entry now uses one consistent card layout: a type icon on the left — its tile tinted with the entry's category colour, so the list colour-codes your life-areas at a glance — the entry's own content as the most prominent line, and a friendly relative date ("Mar 15, 2024") instead of a technical timestamp. Health, workout, measurement and survey entries are shown with human-readable names and compact value chips — e.g. "Systolic Blood Pressure · 122 mmHg" and "Running · 32 min · 413 kcal · 5.6 km" — rather than raw data; checklists show their completion progress with a bar, and habit completions show a clear status ("Completed"). The logbook filter was rebuilt to match the Tasks page: each entry type is a labelled pill carrying its own icon, the starred / flagged / private toggles are clearer multi-select pills, and category pills show their colour. - The Habits tab was reworked to match the rest of the app. It now uses the same calm, light-on-canvas surfaces as Tasks and Time Analysis instead of the old near-black background, with a Time-Analysis-style header. A new summary card shows, at a glance, how many habits you've done today ("2 / 5"), how many are left ("3 to go"), and your current streaks. Your overall consistency now reads as a scrollable heatmap — a calendar of week columns, month-labelled along the top, where each day is shaded by how many of your active habits you completed that day, the way a contributions graph works. It spans the full width of the window and scrolls back through months and years with today pinned to the right edge; quiet (not-yet-done) days stay a plain neutral rather than looking like failures, and every cell announces its date and count to screen readers. The habit list itself is now lean and action-focused: each row shows its own streak ("🔥 23"), a clear "＋" button to log it in one tap (turning into a green check once done), and you can still swipe right to mark it done or left to mark it missed (with an instant confirmation). The recording dialog was rebuilt around a Success / Skip / Missed segmented picker with one Record button, and a linked dashboard (for example an exercise habit's charts) now sits cleanly above the form. The completion-rate chart now opens with a headline — your current 7-day average, whether it's up or down versus last week, and how many points it is from your goal — above a smooth 7-day-average trend line that climbs through a shaded "on track" band, with the raw daily rates as light dots behind it and a quiet, encouraging pointer below to the one habit most worth a focus ("kept 5 of 30"). You can switch the chart between 14-, 30-, and 90-day windows to see anything from a fortnight to a full quarter of progress. Logging a habit is now a small, staged celebration: the check pops in, a soft accent glow blooms on the row, a burst of sparks flies from the check, its streak flame pulses and ticks up, the done count counts up and the progress bar eases forward, and finishing the last habit of the day lights the summary card with an all-done glow. The whole celebration respects the system "reduce motion" setting and announces completion to screen readers. ## [0.9.1026] ### Added - A new "Code" filter pill above a task's linked entries lets you show or hide coding prompts, alongside the existing Timer, Audio, and Images toggles. ### Changed - Coding prompts generated from an audio note now attach to the parent task — the same way cover art does — instead of hanging off the individual audio entry. Because each coding prompt becomes part of the task, follow-up prompts can build on the ones that came before, so you can record a new note asking to continue from an earlier prompt and it has that context. - Editing an entry's start and end time is simpler: you set the date once and the start/end times separately, instead of picking the full date on both pickers. One date wheel sits above paired start/end time wheels, with a live duration. If the end time is earlier than the start, it automatically rolls to the next day and shows a "+1 day" badge — so logging something past midnight needs no extra steps. For entries that span different days, an "Ends on another day" toggle reveals a separate end date. - Creating a project now opens a compact overlay — a bottom sheet on phones, a centered dialog on desktop — instead of taking over the whole screen. The fields are unchanged (title, category, target date); Cancel/Create sit inside the overlay, and the new project shows up in the list as soon as you save. - When the image provider turns down a cover art request, the modal now says the provider rejected it and shows the provider's own reason verbatim (for example "PROHIBITED_CONTENT") instead of a generic "Failed to generate image" — so it's clear the block is the AI provider's content policy, not Lotti, and you can reword the request and try again. - Every button across the app now uses one unified style, and modal action bars share one layout where the confirming action is the wide, dominant button and the cancel/secondary actions stay compact beside it. Disabled buttons now clearly read as inactive — a flat, greyed-out pill — instead of a faded version of their active colour. This covers every previously divergent surface, including the delete-confirmation dialog, the date/time and task due-date/estimate pickers, the journal entry date-range editor, the Matrix sync screens (setup, bundle import, device verification, and the maintenance modals), the AI settings screens (error/retry states, provider setup, and delete dialogs), health import, the agent editors, and habit completion. ### Added - On Linux and Windows you can now import images from a native file dialog via the "+" create menu → "Import image". Previously image import was only offered on macOS and mobile. ### Fixed - On phones, save and action buttons on several settings detail screens were hidden behind the bottom navigation bar, so the action couldn't be tapped — most visibly the Save button on the AI provider setup screen (for example after pasting a Gemini API key). These screens now keep the whole bottom edge: the AI provider connect/edit form, the agent template and soul editors, the sync conflict resolver, and the agent evolution chat input all stay clear of the navigation bar, so their actions are reachable. - Dragging an image or audio file onto a task or journal entry now works on Linux. Two drag-and-drop plugins were both claiming the drop and the wrong one silently swallowed every drop; Lotti now uses a single drag-and-drop system on every platform. ## [0.9.1025] ### Added - Experimental on-device text-to-speech is now available on Linux too (in addition to macOS and iOS): with the "Enable local AI summary playback" flag on, the AI summary card's play button reads a task's TL;DR aloud — or the full report when expanded — synthesized locally with the Supertonic model, no cloud. Still off by default; the model (~400 MB) downloads once on first use. ### Changed - On phones, the Sync settings list now follows the same order and layout as the desktop settings tree, so the two surfaces no longer disagree. The provisioned-sync QR pairing entry is the first row in the list (Provisioned Sync, This device, Backfill, Stats, Outbox, Conflicts, Matrix maintenance); the Sync rows keep their teal icons, and the Outbox row shows the teal postbox with its live pending-sync count badge — now on desktop too, not just mobile. ## [0.9.1024] ### Changed - Category selection is now one consistent picker across the app. Assigning a category (journal entries, tasks, projects, habits, dashboards, time blocks) opens the same searchable sheet and applies as soon as you tap a category. Filtering by category in the habits and dashboards views now uses that same sheet and applies when you tap Apply, rather than each tap taking effect immediately. - A task's labels now open in that same picker: a searchable, multi-select sheet scoped to the task's category (plus any labels already assigned), with an Apply button and the option to create a new label inline. - Search fields no longer sit in a darker box than their surroundings in dark mode. Every search field (tasks, settings, the category and label pickers) is now a subtly raised field that matches its surface instead of a sunken dark well. ## [0.9.1023] ### Added - Experimental on-device text-to-speech (macOS and iOS): a play button on the AI summary card reads the task's TL;DR aloud — or the full report when it's expanded — synthesized locally with the Supertonic model, no cloud. Pick the voice, model, and reading speed under the new Settings → Speech page. Off by default; opt in via the "Enable local AI summary playback" flag. The model (~400 MB) downloads once on first use. ## [0.9.1022] ### Changed - Settings (mobile): the Settings menu was rebuilt to match the desktop tree. The scroll-shrinking title header is gone — a fixed header sits above a clean drill-down list whose rows (icon, title, and a full, no longer truncated description) are the same as the desktop sidebar, so phone and desktop now share the same navigation and drill-down behavior. Tapping a section either drills in (Definitions, Advanced) or opens it directly (AI, Agents, Sync, Theming); back returns one level instead of jumping to the top. Rows grow to stay legible at large system text sizes. - Settings: every settings page now wears one consistent header. The scroll-shrinking title is removed throughout (not just on the menu), and the section lists, editors, sync, and AI pages all use the same fixed title typography and the same back button in the same place — instead of each page carrying its own header style. The left gutter of the rows was also tightened on both phone and desktop. - Insights dashboards: the chart cards were rebuilt on the app design system. Charts no longer sit in near-black boxes in dark mode; value axes use rounded, evenly-spaced ticks with legible, never-clipped labels (and the top value is now labelled); the date axis ticks at the same cadence across every chart and its edge labels no longer clip; bars and lines reach the same right edge and use one higher-contrast series colour (the Steps chart keeps its bars colour-coded by your daily step goal); the blood-pressure chart shows a systolic/diastolic legend with its lines untinted and a calmer 80/100/120 value axis; every card carries a single caption — its unit, the measurable's description, or its aggregation ("Daily total", "bpm", "kcal", "mmHg", …) — instead of the old "[dailySum]"; the weight chart is titled "Weight" rather than "Weight vs. Body Mass Index"; and each chart shows an explicit "No data in this range" state instead of a flat zero line. The add-value (+) button on measurement and survey charts now stays inside the card on desktop instead of being pushed off-screen. The time-span selector now uses the shared design-system segmented control (the same one as the Daily OS plan switch and Time Analysis), the habit card matches the other chart cards (no more near-black box), and the chart cards are inset to align with the dashboard title instead of hugging the screen edges. The dashboard header is now a stable bar — the title and the time-span picker sit together and no longer shrink or scroll away — with a link beside the picker to edit the dashboard's definition (opening in a side panel on desktop, full-screen on mobile). - Measurement entry: logging a value (the + on a dashboard chart) was rebuilt on the design system. The value is a large hero field with the unit inline (e.g. "750 ml"); your most-used amounts appear as one-tap Quick-add chips right below it; the observed-at time is a clearly-editable row; the comment is an optional note; and the Save button stays in place, enabling once the value is a valid number (Enter saves too). - Habit completion: tapping a habit's row while viewing its dashboard no longer re-opens that same dashboard behind the completion sheet — the sheet now shows just the completion form. Opening the habit from the Habits tab still shows the linked dashboard for context. ## [0.9.1020] ### Added - Time Analysis (desktop): a Compare toggle puts the current period side-by-side with the one immediately before it. The headline tile shows the change versus the previous period and the per-category table gains Previous and Change (Δ%) columns. The comparison is honest about in-progress periods — it weighs the same elapsed days against last period's same days (labelled "same days"), not a half-finished week against a whole one — and the change arrows stay neutral for sub-1% noise. The chart itself stays a single, clean series and points to the table ("Comparison shown in the table below") rather than crowding in a ghost reference bar. ### Changed - All settings definition pages (categories, labels, habits, measurables, dashboards) now share one design language. Every list page has the same header, search field, grouped rows, floating create button, and localized empty/no-match/error states; every editor shares the same detail layout — a sticky bottom action bar floating on blurred glass with Save/Create (enabled once something actually changed), Cancel, and Delete with confirmation, Cmd/Ctrl+S keyboard save, design-system text fields and toggle rows grouped into titled section cards, and content that stays aligned with the page title at every window size. Editor titles read "Edit habit"/"Edit category"/… consistently, and stray trailing colons on field labels ("Priority:", "Dashboard name:") are gone. - Creating a category now opens the same full create page as every other definition type instead of a popup. It asks only for name, color, and icon, then lands in the new category's editor where everything else (privacy, language, AI defaults) is configured — no more disabled placeholder switches on the create form. - The Projects section was removed from category settings; projects have their own top-level page now. - On desktop, habit editing now opens inline in the settings split pane like the other definition types, instead of taking over the window. - Dashboard editor: save-and-copy moved to a header action next to the title, and saving with an invalid form no longer navigates away while silently discarding the changes. - Time Analysis (desktop): the wide date-range row — preset chips plus a custom-range dialog — is replaced by a compact period stepper. Pick a granularity (Day / Week / Month / Quarter / Year) and browse to the previous or next period with the ‹ › arrows; the dashboard updates live as you step, with no dialog. Weeks follow your device region — they start on Monday across most of Europe and on Sunday in the US — both in the stepper's week periods and in the jump-to-date calendar grid. MTD and YTD shortcut pills stay on the row: one click jumps to the current month-to-date or year-to-date (through today, not the full calendar period), and with Compare on they line up against the same elapsed days of the previous month or year. - Time Analysis (desktop): tap the period label to open a calendar and jump straight to any date's period; the dashboard updates live behind the sheet as you pick. ## [0.9.1019] ### Changed - The Daily OS day-planning conversation was redesigned around a stable voice orb: the orb now stays put directly above the action bar in every state while the live transcript grows upward above it (no more layout jumping under your finger mid-recording), the orb is smaller and reacts to presses with a visible ripple and a press-scale animation, and while recording its core breathes with your voice — swelling with the level and keeping a slow idle pulse between words — under a stop glyph; after capture the orb steps back so the Review action carries the weight. Every action lives on the sticky bottom bar in every state — including a Done pill while recording and a Cancel while transcribing — and the per-state headline narrates what is happening ("I'm listening.", "Writing that down…", "Does this look right?"). - On desktop, day planning now opens in a full-height side panel on the right (45% of the window) instead of a centered dialog, so the day stays visible while you talk; panel buttons size to their content instead of stretching across the panel. - The Drafting wait replaced its empty shimmer boxes with the AI thinking animation over a rotating narration of what the agent is doing ("Reading your check-in…", "Placing deep work first…"), with yesterday's learning cards as real content to read while waiting. - The Refine step was rebuilt on the same layout as Capture: it now shows the current plan you are about to change, streams your words above the orb, and offers Revert / Looks good on the bottom bar. - Day-planning copy and typography polish: a "Review" CTA instead of the jargon "Reconcile", a real headline on the review step, typographic apostrophes, and large accessibility text sizes no longer overflow the capture sheet on small phones. - The Daily OS day view now reads as paint-by-numbers: planned blocks are a faint sketch of their category color (dashed while still drafted) and recorded sessions carry the full, alive color — doing fills the day in. The timeline opens centered on the now-line, which spans both the plan and actual lanes, and short blocks no longer clip text mid-letter. - The day's agenda is much calmer: finished items collapse into one-line receipt rows so what's next owns the screen, metadata became quiet icon+text instead of a row of pills, "in progress" is no longer painted amber, and the capacity ring is now a thin stacked category ring that matches the legend, with the remaining time in the center. - The Daily OS day page gained an Apple-style large-title header — a quiet back chevron on its own line, "Your day" with a weekday-first date overline, and the toggle, nudge chip, and cards all aligned to one left rail. First-run gesture and lock-in coaching now retire after you use them once instead of lingering, and content dissolves softly at the fold under the glass bar instead of being cut off. The capacity card leads with the time you have left (e.g. "15m left"), demotes the committed/capacity detail to a supporting line, and turns the category legend into a right-aligned receipt table; every concrete date across the Daily OS now carries its weekday. - Learnings the planner wants you to confirm now surface right on the day page as a quiet "things I noticed — review" chip (plus a permanent "What I've learned" entry in the day menu) on both mobile and desktop, instead of being reachable only from the evening wrap-up screen. - Light theme contrast: secondary "low emphasis" text and the teal accent were too faint to read for low-vision users (about 2.2:1 against the background). Light-mode secondary text is darker now, and the light teal accent chain is deeper, so captions, hints, and white-on-teal buttons all meet accessibility contrast. Dark mode is unchanged. ## [0.9.1018] ### Added - The Daily OS planner now sees the last week in context when drafting a day: planned-vs-recorded time per category, missed sessions by name, and upcoming plans plus deadlines for the next five days. It also keeps a short note on how each day actually went, written as the day closes, and reads those notes back later — so after an overworked stretch it plans a gentler day instead of piling on. Sustainability over throughput. ### Changed - Mobile: the bottom navigation bar no longer crams every destination into one row with shrinking labels. It fills with as many destinations as comfortably fit your window and text size: on narrow screens it shows Tasks, DailyOS (when enabled), and Logbook plus a "More" slot, with Settings and the remaining flag-gated destinations (Projects, Habits, Insights) in a More bottom sheet — that's also where newly toggled pages appear, and the Settings row shows its pending-sync count on the right like the desktop sidebar does. While a sheet destination is on screen, the More slot takes its name and the active highlight so you can always see where you are. As the window gets wider, destinations move out of the sheet into their own slots one by one — always in the same order and positions — until everything fits and the More slot disappears entirely; a larger system font widens the labels and brings the overflow back sooner. Daily OS also moved directly after Tasks across mobile and the desktop sidebar. The bar docks flush against the bottom screen edge — full width, no floating gap — with the home-indicator inset absorbed into the bar surface, and its top corners now use the gentler card radius (matched by the settings header's bottom corners). - Mobile: opening a settings entity-definition editor — the detail or create page of a category, habit, label, dashboard, or measurable, or a project editor — now slides the bottom navigation bar smoothly below the screen edge, giving the editor the full height, and slides it back in when you leave; the list pages keep the bar in place. A running timer or audio recording indicator stays visible throughout, dropping to the screen's bottom edge while the bar is away. ## [0.9.1017] ### Added - New desktop "Time Analysis" dashboard under Daily OS, opened from a sidebar entry beneath the month calendar onto the full screen: see where your time went per category and per day, with quick range presets (1d, 7d, 30d, MTD, YTD, last month) plus a custom date range, a daily/cumulative chart toggle, headline totals, and a precise per-category table with share and daily averages. Optional "focus categories" split the headline into focus vs. other time. Range switching is instantaneous even with years of entries, and private entries stay hidden unless the private flag is on. ### Changed - The desktop sidebar's running-timer card now stays visible the whole time a timer is running. Previously it hid itself while you were viewing the running task and could also disappear after you navigated away from the Tasks tab, so the elapsed-time readout and the quick jump-back button were easy to lose. The card now persists in the sidebar everywhere — including while the same task is open in the main view, where it sits alongside the task's own running indicator. - Task and project lists stay snappier under sync: a task's project is now resolved from a denormalized, indexed column (and batched across rows instead of one query per task), and the projects overview plus the saved-filter sidebar counts now coalesce rapid update notifications into a single refresh instead of recomputing on every burst. ### Fixed - Time tracked on entries created before mid-2024 could show up as "Uncategorized" in time-by-category views (including the Daily OS time history): the denormalized category column is now backfilled once from the stored entry data. - Daily OS Next: tapping a recorded session on the Day timeline now reopens its task scrolled to — and briefly highlighting — that exact entry again, instead of dropping you at the top of the task. - Daily OS Next: the sidebar month calendar now starts the week on your device region's first weekday — Monday across most of Europe, Sunday in the US — read from the OS region (natively on macOS) rather than the app language, instead of always starting on Sunday. ## [0.9.1016] ### Added - Categories have a new "Day planning" switch. Day-plan category selection is strictly opt-in: only categories with the switch turned on are offered in the day plan — nothing is available by default. - Projects have a new "Monitoring" status for projects that are not closed yet but have no time actively scheduled — they are only touched when something comes up before they can be declared done. - Task agents now maintain their own day-planning attention requests during normal wakes, resolving stale requests when tasks are done or no longer need scheduled time. - While the Daily OS tab is active, its desktop sidebar entry expands into a month calendar: today is highlighted, days with a Daily OS plan carry a dot, and tapping a day jumps straight to it. - Daily OS Next: a day with tracked time but no plan now opens on the Day timeline so recorded sessions are visible without drafting a plan first, with an honest "No plan yet" capacity strip and a "Speak a check-in" button into Capture. - Daily OS Next: a bounded "Today so far" card shows the day's recorded sessions on the Capture screen and on the empty Agenda tab, collapsing older sessions behind an "N earlier sessions" expander. - Daily OS Next: agenda items and day blocks now show whether they are backed by a task (blue link badge that opens the task) or are standalone time blocks, and standalone titles can be renamed inline (click to edit, Enter or blur saves, Esc cancels). ### Changed - Daily OS Next visuals now follow the latest design handoff: a capacity donut replaces the linear meter on the Agenda and Commit surfaces, recorded sessions render in a neutral "tracked" treatment with mono time ranges, drafted blocks carry a dashed outline, the commit sign-off uses the "Final step / Make it yours." composition with a smaller single-word hold button, and typography across the feature is calmer (lighter overlines and headers). ### Fixed - Slow-query logging now starts from persisted logging settings during startup, so enabled `slow_queries` and `super_slow_queries` files capture boot-time database work instead of waiting for the next settings toggle. - Daily OS Next day planning now forces required draft tool calls through Gemini and Mistral providers and keeps fast completed drafts from being dropped while learning cards load. ## [0.9.1015] ### Added - The linked entries "Filter & Sort" modal now has a "Show flagged entries only" toggle that narrows the list below a task to flagged entries; the list updates immediately when the toggle or an entry's flag changes. - New "Agent fork healing" flag in Settings → Flags (off by default): when enabled, agents whose histories diverged through multi-device use merge them back together at the next wake. ### Changed - Agent memory compaction now runs by default for task, project, and day agents, and the Settings flag that controlled it has been removed. - Config flag changes from Settings now sync to the user's other devices when the flag is flipped, without broadcasting default startup flag seeding. ### Fixed - Label usage counts in the labels list no longer include deleted journal entries, and entries marked private only count while the "Show private entries" flag is enabled. - Task-agent suggestions now suppress verbatim duplicate proposals even when their underlying tool arguments differ, so identical checklist suggestions appear only once. - Daily OS Next capture wakes now retry with a forced parse step when the first model response skips parsing, so Reconcile no longer stays stuck with an empty "Here's what I heard" column. ### Fixed - Saving an event now clears the unsaved-changes state, so the save button no longer stays visible after a successful save. - Buttons that customize the primary button style (e.g. the date picker's Done button) now apply their custom disabled color instead of silently falling back to the default. ## [0.9.1014] ### Added - New "Agent memory compaction" flag in Settings → Flags (off by default): when enabled, task agents summarize older task-log entries into a rolling memory using their own model, keeping the wake context compact while recent entries stay verbatim. ### Changed - Task agents now update their report only when its content would materially change, instead of re-publishing it on every wake — reports churn less and wakes finish faster when nothing report-worthy happened. - Task agents can now propose archiving a duplicate or obsolete checklist item; the proposal appears in the usual confirmation flow as "Archive: …" and archiving stays reversible. - The agent conversation view now shows the exact system prompt each wake ran with as an expandable "System Prompt" row (stored once per distinct prompt), orders each wake's rows in conversation order, and labels bookkeeping rows as "Milestone"/"Retraction" instead of a generic "System". ### Fixed - Task agents now fall back to a pending title-change proposal when the immediate initial-title write is blocked by category policy, instead of repeatedly reporting a policy-denied tool error. ## [0.9.1012] ### Changed - Task detail pages now show running AI activity as subtle shader-based decoder bars integrated into the sticky action bar, replacing the separate Siri-wave strip above it. - Daily OS Next voice capture now wraps the record button with the dBFS-driven tension-loop shader while listening, and shows reviewed transcripts in a calmer fixed-height review panel. ### Fixed - Task due-date urgency now uses calendar-day comparisons, so overdue and due-today styling stays correct across daylight-saving transitions. - Plain-text journal/comment content with quotes, backslashes, or line breaks now stores valid rich-text editor data instead of interpolated JSON fragments. - CSS hex color parsing now rejects malformed strings with embedded valid colors, keeping fallback colors predictable for invalid input. - AI chat streaming now preserves split reasoning close markers and final soft line breaks when providers deliver responses one character or token at a time. ## [0.9.1011] ### Added - Settings → Advanced → Logging now has an individual on/off switch for every logging domain (sync, AI, chat, speech, persistence, database, agent runtime, agent workflow, tasks, labels, health, habits, location, screenshots, calendar, navigation, theming, notifications, what's new, settings, ratings, Daily OS, general), replacing the previous three fixed toggles. Sync stays off by default; the rest are on while the master logging switch is enabled. Errors are always logged regardless of these switches. - Errors are now mirrored into two daily files for quick inspection: `error-YYYY-MM-DD.log` with full detail, and a PII-safe `error-safe-YYYY-MM-DD.log` that records the error type only (never the raw exception text), so the safe file can be shared without leaking content. Under normal operation both stay empty. ### Changed - Backfill sync now distinguishes intentionally burned vector-clock counters (benign voided numbers in the monotonic sequence) from genuinely unresolvable entries, with a separate "Burned" count in the Backfill settings screen, so newly burned counters are no longer reported as data loss. Entries already marked unresolvable keep that label until they settle into the correct bucket as peers are re-asked. - Daily OS Next refine-by-voice now opens as a responsive Wolt modal over the current plan surface, and keeps the reviewed transcript visible with inline feedback when a proposal fails or produces no usable changes. - Daily OS Next measures the date controls, Agenda/Day switcher, and header actions before placing them, keeping the switcher inline whenever it fits and moving it below only on truly constrained widths. ## [0.9.1010] ### Added - AI popup skill runs can override the Gemini thinking effort for a single invocation after choosing the model, so one saved model row can still be used with different thinking budgets per use case. ### Changed - Daily OS Next now personalizes the greeting from Settings > Advanced > About, lets typed capture start from the idle screen, filters processing categories, wraps mobile task titles, compares planned and actual timelines with a shared scroll/zoom surface, folds idle midnight-to-midnight day gaps with a 24-hour time rail, shows recorded journal time in the Actual lane, keeps task-backed agenda titles in sync, refreshes without whole-page loading flashes, and avoids surfacing weeks-old overdue tasks in daily proposals. - Daily OS Next agenda rows now use a cleaner mobile-first layout with task cover art thumbnails, tighter metadata pills, token-backed progress, and task-backed Day timeline blocks and task-report links that open the linked task inside the app. - Daily OS Next drafting now carries approved new capture items into planning, counts buffers as scheduled time while excluding dropped blocks, and labels due/overdue work relative to the selected plan date. - Sync diagnostics and backfill stats put less load on the database, reducing sync-related stalls and improving responsiveness on busy machines. ### Fixed - Daily OS Next capture errors now use localized copy instead of raw controller strings. - Mobile: Daily OS reconcile controls now reserve the bottom navigation hit area, and the bottom navigation labels have additional breathing room on platforms without a home-indicator inset. - Improved sync reliability: fewer unresolvable gaps and more dependable delivery of local edits across devices. - Matrix sync now recovers automatically from stale empty downloads, so media and JSON attachments arrive more reliably without manual retries. - Agent task suggestions stay stable while the agent is working: they no longer flash empty or reshuffle as the agent revises them, and acting on one suggested checklist item no longer makes the others disappear. ### Changed - Sticky glass footers in Tasks and Daily OS now use a token-backed scrim so action controls stay readable when bright screenshots or light-mode content scroll underneath them. - Tasks category filter and journal entry-type filter chips have a small extra nudge of space between the chip border and the label so the text no longer sits flush against the leading edge. ## [0.9.1008] ### Added - Gemini model settings now include a configurable thinking mode with minimal, low, medium, and high options; existing models default to low for faster inference unless changed. - Active audio recordings now show a prominent desktop sidebar card above the running timer, with a larger speech-weighted dBFS-reactive recording orb, red signal-reactive frame, subtle red shadow, elapsed time, and one-tap stop control so runaway microphone sessions are harder to miss. - Audio and image entry Actions menus on desktop can now reveal the backing file in Finder, File Explorer, or Files. - Daily OS day-agent capture/reconcile tooling now persists captured text, parses it into task decisions, links it to existing tasks, and applies triage actions through the agent backend. ### Changed - Tasks navigation no longer shows the global in-progress task count badge in the sidebar or bottom navigation; saved filters remain the focused way to monitor task subsets. - Daily OS Next day planning now lets you edit capture transcripts before reconciliation, creates real tasks from new capture phrases, opens task-backed agenda rows, rejects new same-day draft blocks that start in the past, amends approved plans through per-change tracked diff approvals, replaces the cramped capacity donut with a linear meter, lets you review and edit refine transcripts before they reshape the plan, uses real audio transcription on the refine screen instead of a scripted sample diff, and prefers Mistral cloud speech recognition over MLX when both are configured. - Gemini-backed speech recognition, Daily OS day planning, and task-agent wakes now respond faster on interactive turnaround. - Agent batch lookups for latest per-agent reports and states now use ranked active-row indexes that avoid SQLite temp sorts during large project-agent wake context builds. ### Fixed - Sync now preserves local sequence mappings when recovering burnt vector-clock counters, reducing cases where peers can get stuck on unresolvable gaps. - Task-agent suggestions no longer disappear wholesale while an agent wake is running, and stale running-timer update proposals are automatically retracted if the timer stops before you accept them. - Agent workflow diagnostics no longer write suggestion argument values or raw exception text into runtime log files. - The Create Category modal's colour picker no longer overflows on desktop: the hue slider used to be clipped off the right edge of the modal, leaving the colour ungrabbable. The picker now fits inside the modal width on every platform, with the saturation square clamped to the design-system size. - The task detail action bar no longer shifts or leaves faint ghost shadows behind the Track time, audio, checklist, and more-action buttons when time recording changes state. ## [0.9.1005] ### Changed - AI popup menu now opens a model-override picker when you tap an Analyze Image skill, mirroring the existing Transcribe flow — the bottom sheet lists every image-capable model you've configured so you can route one specific photo to a different model without editing the inference profile. Single-model setups short-circuit to preserve the one-tap flow; a stale override falls back to the profile slot. - Upgraded Flutter to 3.44.0 and aligned the Flatpak Flathub manifest template with the repository FVM toolchain pin, so source builds submitted to Flathub use the same Flutter SDK as local and CI builds. ## [0.9.1003] ### Changed - AI popup menu now opens a model-override picker when you tap Transcribe on a voice note. The bottom sheet lists every speech-capable model you've configured — the profile's default marked with a "(default)" badge and a check at the top, alternatives below — so you can route one specific transcription to a different model (e.g. cloud Mistral instead of local Voxtral) without changing the entire inference profile. When only one speech-capable model is configured the picker short-circuits, preserving the one-tap flow for the common case. A stale override (model deleted between picker and run) falls back to the profile's slot rather than stranding the user. - Inference-profile model slot picker now has a search field at the top. Substring match runs against the model display name, the wire-level `providerModelId`, and the resolved provider label — so typing "Google" surfaces every Gemini-owned model even when the row text doesn't start with the provider name. When the query matches nothing, the picker shows the existing "No matches" empty state instead of leaving the list region blank. Clearing the query restores the full slot-filtered list. - AI Settings — the legacy "Select Provider Type" sheet has been folded into the modernized "Set up AI features" picker. Power users who previously dismissed the FTUE prompt, and the in-form provider-type switcher inside the provider edit page, now open the same modern modal in a non-FTUE mode that hides the subtitle, footer hint, and "Don't show again" button while surfacing every `InferenceProviderType` — including OpenAI-compatible, OpenRouter, Nebius, Mistral, and Whisper — alongside the curated branded tiles. One widget, one visual language, no provider type unreachable. - "Add Model" from inside a provider's detail page now preselects that provider on the new-model form, so the user no longer has to re-pick the provider they just came from. The top-level "+ Add model" FAB still opens a blank form where the user picks the provider manually. Edit-mode (existing model) is unaffected — the form continues to honor the model's stored `inferenceProviderId` and ignores any stray preselection arg. - MLX Audio is now macOS-only. The iOS Runner no longer links `mlx-swift`, `mlx-audio-swift`, or `swift-huggingface` and no longer registers the `MlxAudio` plugin, which strips the on-device STT runtime that triggered immediate OOM with the 1.7B Qwen3-ASR model on iPhone hardware and shrinks the iOS bundle. iOS now reports MLX models as `unsupported` the same way Android and Linux already did. Audio recorded on iOS still reaches MLX through the existing synced-audio auto-trigger flow on a paired macOS desktop — the sync-node capability probe already gates the `mlxAudio` capability on `Platform.isMacOS`, so deferred inference continues to work unchanged. - Category pickers now share the same look across the app. The task filter's Category sub-modal gains a search bar at the top (matching the Settings → Categories Definition page), and both the filter sub-modal and the category modal opened from the task definition page now anchor their action button inside a `DesignSystemGlassStrip` footer, so list rows scroll behind the blurred bar instead of being pushed off-screen. The task definition modal also swaps its `LottiSearchBar` (gradient-filled in dark mode) for the flat, border-only `DesignSystemSearch` used by the settings page. - Desktop slow-query hot paths now avoid several major sources of UI stalls: sync outbox claims and sync-sequence watermark reads use SQL shapes that match their partial indexes, editor draft restore has a composite draft lookup index, agent active-row reads gain dedicated partial indexes, and the agent/task summary paths resolve templates, active soul documents, task-agent links, and latest reports in batches instead of issuing per-row link and entity lookups. ### Added - Synced-notifications bell in the Tasks and Projects tab headers. The icon flips from outlined to filled and shows an unread count badge (capped at `9+`) whenever there are unseen alerts, driven by a new Riverpod provider that listens on `UpdateNotifications.updateStream` so the badge stays in sync with the database. Tapping the bell opens an anchored popover that lists the pending alerts: each row shows the title and body, tapping deep-links into the linked task and marks the alert acted-on, and the dismiss icon (or a long press) retracts the alert. All entry points stay behind the existing `enable_synced_alerts` master switch and reuse `NotificationRepository`, so every change syncs to other devices through the existing Matrix sync pipeline. ### Fixed - Sync outbox polling now uses priority-first send-order paths for pending and expired in-flight rows, avoiding temp B-tree sorts in the hot dequeue loop while letting journal/link updates drain ahead of older agent work. Sequence-log watermarks are now persisted per host after a lazy first read, so hot gap detection reads one primary-key row instead of recomputing the contiguous prefix with a `ROW_NUMBER` scan over the sequence log. Inbound stranded-room pruning now also has a literal-status partial index over active queue rows, and journal sync diagnostics report rows written as a count instead of leaking SQLite rowids. - Incoming agent entity/link sync now respects vector-clock dominance before overwriting local agent state, while still recording skipped stale/equal payloads in the sequence log. - Agent report and task-link bulk lookups now chunk large `IN` lists, so high-volume sync payload collection no longer risks SQLite host-variable limit failures when thousands of agent/task ids are considered at once. - Habit completion reads now enforce last-write-wins per habit/day, so repeated success/skip/fail updates on the same date resolve to the newest persisted write instead of depending on database row order. - Flatpak release generation now carries exact ObjectBox 5.3.2 offline source hashes and a CRLF-tolerant CMake patch, plus a PR-safe validation job that dry-runs Flatpak foreign dependency patches against the locked Pub packages after dependency updates. - Agent wakes no longer abort when synced duplicate model rows leave the first `providerModelId` match pointing at a deleted or unusable provider. Provider resolution now skips stale candidates and prefers the usable provider with the known matching provider type, while known-model backfill and FTUE setup stop seeding duplicate rows for the same provider-native model ID unless the existing row is orphaned and needs repair. - Task-agent proposal ledgers no longer surface retired suggestions as open work. Consolidating multiple pending change sets now marks pending items in the retired source sets as `retracted`, and the ledger reader filters historical resolved-parent/pending-child snapshots plus stale rejection or retraction decisions before the prompt and AI summary card see them. `retract_suggestions` also treats fingerprints already present in the resolved ledger as `not_open` rather than `not_found`, avoiding repeated model retraction loops. - Task-agent suggestion notifications now enforce one active bell row per task. Fresh agent waves still get durable change-set-seeded rows, but the repository serializes same-task notification mutations, retracts older open rows before surfacing the latest pending count, and bulk-clears every open suggestion row when the user or agent resolves the proposals. Tapping a suggestion notification now also focuses the task-agent proposals section, including when that task detail is already open. ## [0.9.1002] ### Added - Synced notifications data layer: a separate `notifications.sqlite` store, Freezed notification entities, Matrix sync payloads for full alerts and state updates, backfill coverage, and macOS/iOS scheduler plumbing behind the new `enable_synced_alerts` flag. - Automatic AI inference pickup on capable desktops when a mobile device syncs an audio entry. Each node broadcasts a sync node profile (display name, platform, detected capabilities — MLX Audio, Ollama, Voxtral, Whisper) over Matrix and maintains a local directory of peers. Inference profiles gain an optional pinned device: when a mobile-recorded `JournalAudio` arrives via sync and its responsible profile is pinned to the local host (and the profile is fully local — every populated model slot resolves to MLX Audio, Ollama, Voxtral, or Whisper), the desktop runs the existing transcription skill and nudges the wake cycle — no manual transcription, no cloud routing. Pinned host offline = entry stays untranscribed; the user can re-pin. New AI Settings → "This device" page edits the local display name, shows detected capabilities, and lists known peer nodes. The inference profile edit page exposes the Pinned device selector, filtered to peers that advertise the capabilities the profile actually needs. - Live API-key verification on the AI provider connect form. As soon as you stop typing in the API key or Base URL field, the form checks the credentials against the provider's API and shows a status strip below the Base URL with one of three outcomes: a spinner while the check runs, a green confirmation with the number of models the key can see and the response time, or a warning row explaining what went wrong (invalid key, network error, timeout, unreachable host) with a Retry button. Covers Gemini, OpenAI, Anthropic, Mistral, Alibaba, OpenRouter, Nebius, and Ollama out of the box — no more "save the form, hope the key works, find out the next time you try to chat". - Alibaba Cloud (Qwen) promoted to a first-class FTUE option. The pick-provider modal now carries a fifth tile for Alibaba with the NEW badge, sitting between Anthropic and Ollama, and the zero-providers state card adds a matching fifth chip in the same position. Alibaba also gets its own brand-orange accent token (light `#E0762A` / dark `#FFA868`) and a tagline ("Qwen models · multimodal · long context") so its tile and cards no longer fall back to the neutral interactive accent. Translations land in all six locales (informal tone). - Voxtral (local) tile in the AI Settings pick-provider modal, added as the sixth tile (after Ollama) with the same `DESKTOP ONLY` badge. Picking it routes to the connect form preselected to `InferenceProviderType.voxtral` with the default base URL (`http://localhost:11344`) prefilled and no API key required. Until this release there was no UI affordance to add a Voxtral provider, even though the rest of the runtime (repository, default URL, known models, chip + visual) has supported it for some time. - MLX Audio (local) provider scaffolding for embedded Apple Silicon speech. AI setup now offers MLX Audio, seeds Voxtral, Qwen3-ASR, Parakeet, and Qwen3-TTS model rows, prompts for model installation, and shows native download progress on both model overview and provider detail surfaces. Post-recording transcription routes through a guarded Swift channel. AI-summary playback remains wired through the same channel on macOS but is hidden behind the default-off `enable_ai_summary_tts` config flag while local TTS is evaluated. The iOS target does not link `MLXAudioTTS` yet because its Moss TTS target is not archive-safe on iOS; iOS `speakText` returns unsupported while STT remains linked. Intel macOS and builds without the MLX Audio Swift SDK keep compiling and report the feature as unsupported instead of loading Apple Silicon-only libraries. - MLX Audio now also offers Qwen3-ASR 1.7B 4-bit and 8-bit model rows for local post-recording transcription tests on macOS and iOS. First-run MLX Audio setup now lets users choose which local STT model to install, with Qwen3-ASR 1.7B 8-bit preselected because it is materially faster than Voxtral Realtime for post-recording transcription. Batch transcription prefers configured MLX Qwen3-ASR models so speech-dictionary terms can be passed as Qwen prompt context; the realtime code path remains in place but its UI toggle is disabled until live transcription can use comparable biasing. MLX model download status is now held in a shared progress store so overview cards, provider-detail cards, and the modal render the same live percentage and the progress modal can be reopened while a download is running. Speech recognition can also fall back to a configured audio-to-text model directly when no applicable inference profile is available, which keeps local MLX STT usable on mobile without requiring a desktop-only profile. The native bridge now refuses transcription against missing or partial MLX models instead of starting an implicit background download from an inference call, and logs memory-stage diagnostics around model load and generation for crash investigation. ### Changed - AI default-profile seeding is now strictly seed-on-create. User edits to a bundled profile (e.g. swapping the Ollama thinking model) survive restarts; previously the seeder reconciled model slots and flags back to the bundled values on every launch. - AI Settings delete confirmations now use the same design-system toaster that the checklist row uses (warning tone, 5-second countdown bar, Undo action), replacing three custom hand-rolled SnackBars for provider / model / profile / prompt / skill deletions and a fourth for delete failures. The provider cascade list, previously rendered as a custom bullet block inside the SnackBar, now collapses into the toast description ("Also removed 2 models: gpt-4-turbo, claude-3.5-sonnet") so long lists ellipsize naturally. All copy moved into the `aiDeleteToast*` localization keys across English, Czech, German, Spanish, French, and Romanian (informal register). - Profile cards in AI Settings now pin the `⋯` overflow menu to the far right of the card header — matching the provider cards — instead of letting it drift next to longer profile names. - AI Settings Profiles tab Active badge now reflects whether the profile is the winning candidate for at least one configured provider — the same rule the provider detail page already uses for its "Active profile" section. Previously the badge was wired to `profile.isDefault`, which the seeder stamps on every shipped profile, so every seeded card lit up the green badge regardless of whether the user had actually wired the underlying provider. The picker logic moved into a new `lib/features/ai/ui/settings/util/active_profile.dart` so both surfaces consume the same definition; the detail page's Active-profile card also now shows the badge unconditionally (it can no longer fall off the card because of an unrelated `isDefault` flag flip). "Configured" mirrors `AiProviderCardStatus.statusFor` returning `connected`: cloud providers need a non-empty API key (drafts are skipped, even when their model rows survived the API-key clear), and Ollama needs a non-empty base URL plus at least one model row. The badge can no longer light up while the underlying provider card reads "Invalid key" or "Offline". - The "Local Power (Ollama)" seeded profile now uses `qwen3.6:35b-a3b-coding-nvfp4` (35B MoE / ~3B active, NVFP4 quant, 22GB download, 256K context, text-only) as the thinking model, with `qwen3.5:27b` retained as the image-recognition model. The Qwen 3.6 coding MoE is also added to the known Ollama model list so it can be selected from the model picker. - Category tags now pick black or white text based on the background brightness instead of a fixed palette colour, so the label stays legible on dark category colours (e.g. the seeded Ollama charcoal `#0F172A`) and bright ones alike. - AI provider setup-result modal (the post-FTUE "{Provider} is connected" sheet) now ships a single "Start using AI" CTA instead of a secondary "Review setup" button alongside it. The CTA fills the row on mobile-width surfaces and pins to the right with a comfortable cap on desktop / tablet dialog widths, so the one-button footer reads as a deliberate primary action at every breakpoint. ### Fixed - Agent wake-cycle deadlines now survive app restarts as executable work. Persisted `nextWakeAt` rows are restored into the in-memory wake queue, so a 06:00 deferred wake that becomes due while the app is closed runs after the afternoon restart instead of lingering in the sidebar as stale. Completed subscription wakes also stop writing cooldown-only `nextWakeAt` rows when no follow-up job is queued, so the Wake Cycles list only shows work that can still run. - Model IDs in the category Default-inference-profile picker no longer render with extra space between every digit. Replaced the generic `fontFamily: 'monospace'` (which Flutter doesn't ship, so macOS fell back per-glyph) with the bundled-Inconsolata `monoMetaStyle` helper. - Voxtral (local) and Whisper (local) provider cards and detail-page status pills no longer report "Invalid key". The status helper only exempted Ollama; it now uses `ProviderConfig.noApiKeyRequired`, so all three local providers share the same base-URL + model-count gate. - Agent wake-cycle countdowns now switch to `h:mm:ss` once a scheduled wake is more than an hour away, so next-morning wake deadlines no longer appear as giant minute counts like `385:20` in task/project AI report cards. Task-agent propagated updates also use the normal 120-second coalesced wake path instead of being pushed to the next 06:00 digest slot, and live wake titles now update across repeated agent-originated task title changes. - Adding a second provider after tapping "Don't show again" on the FTUE pick-provider modal no longer strands the user on a `genericOpenAi`-prefilled connect form. The dismiss flag (`AI_PICK_PROVIDER_DISMISSED`) was routing every subsequent "+ Add provider" tap to `navigateToCreateProvider(context)` with no `preselectedType`, and `InferenceProviderFormController.build` defaults to `genericOpenAi` in that case — so users who'd set up Gemini first and then wanted Ollama saw an "OpenAI-compatible setup" with no surfaced way to switch types. The handler now routes dismissed users through the legacy `ProviderTypeSelectionModal` (a new `showForResult` entry point resolves with the picked type via a `Completer`) which lists every `InferenceProviderType` including Ollama, Voxtral, Whisper, OpenRouter, Nebius, and the generic OpenAI fallback — so the "start with Gemini, then add Ollama" workflow now works the same regardless of whether the dismiss flag is set. - Speech recognition (and other entry-level AI skills) now run on standalone audio/text/image entries that have no parent task, using the entry category's default inference profile. Previously the AI popup-menu trigger silently no-oped whenever an entry had no linked task, so users with a category configured for Ollama + Voxtral could open the menu and tap Transcribe with no feedback and no transcription. `ProfileAutomationResolver` gains a `resolveForCategory()` path reading `CategoryDefinition`'s `defaultProfileId`, and `triggerSkillProvider` falls back to it when `linkedTaskId` is null. The popup also hides skills whose `contextPolicy` requires the full task (cover art, coding / design / research prompts, image analysis and transcription "in Task Context" variants) for standalone entries, and now applies the same fail-closed rule to `Modality.text` skills — those are only offered on the four text-bearing entity surfaces the popup actually renders on (`JournalEntry`, `JournalAudio`, `Task`, `JournalImage`), matching the existing audio/image-modality branches. ## [0.9.999] ### Added - AI Settings FTUE pick-provider modal. Tapping the "Add provider" FAB on the AI Settings page now opens a centered `AiPickProviderModal` first instead of jumping straight to the empty connect form. The modal carries a `Set up AI features` title, a one-line subtitle, and four big provider tiles (Gemini RECOMMENDED, OpenAI, Anthropic NEW, Ollama DESKTOP ONLY) with radio-style selection. Footer hint pins the reassuring "Your API key is stored locally" copy plus a secondary `Don't show again` and a primary `Continue ▸`. Picking a tile and hitting Continue routes to the connect form preselected to that provider type. Tapping Don't show again writes a settings flag (`AI_PICK_PROVIDER_DISMISSED`) so subsequent FAB taps go straight to the form — the picker never harasses a user who has told us they know the flow. - DRAFT badge on half-configured provider cards and on the provider detail header. A cloud provider with no API key yet (saved via the upcoming "Save as draft" affordance, or one whose key was later cleared) renders a secondary-toned `DRAFT` outlined badge next to the provider icon on `AiProviderCard` and inline with the display name on `_HeaderStrip`. Local providers (Ollama, Whisper, Voxtral) never need a key, so they are never reported as drafts. Surfaces in [[ai-provider-visual]] via the new `isProviderDraft` helper so the cards and the detail page share one source of truth. - `popAiSettingsDetail()` back-affordance shared by all AI Settings detail pages (`AiProviderDetailPage`, `InferenceModelEditPage`, `InferenceProfileForm`). On mobile / pushed stacks it pops; on the desktop master/detail panel where the page sits in the panel slot (not on the Navigator stack) it falls back to beaming `/settings/ai`. Fixes the "back button does nothing" regression on the desktop surface that previously left users stuck on a detail page with no exit. Defensive guard means widget tests using a bare `MaterialApp.home` no longer crash on the GetIt lookup — the helper bails out silently when no NavService is registered. ### Changed - Inference profile form chrome aligned with the rest of the AI Settings v5 pages (model edit + provider detail). Scaffold and AppBar now use `tokens.colors.background.level01` (the AI Settings surface tone) instead of the light theme's `surfaceContainerLowest` / dark theme's `scrim` pair, the AppBar is flat (`elevation: 0`, `scrolledUnderElevation: 0`), the title uses `subtitle1` typography, and the page padding is token-driven (`spacing.step5`). - "Add provider" connect-form chrome redesigned to match the typed three-step flow. The legacy `Add Provider` SliverAppBar title is replaced in create mode by `_CreateModeChrome`: a breadcrumb row (`Settings › AI Settings › Add provider › `, shown on viewports ≥ 720 px), a step indicator (`Choose provider › Connect › Review` with Connect highlighted), and a provider-tinted hero card pulled from `aiProviderVisual` carrying the localised `Connect ` headline and tagline. Below the form the legacy `FormBottomBar` is swapped for an `_AddProviderFooterBar` with three actions: `Back to providers` (back-arrow tertiary), `Save as draft` (secondary, fires save WITHOUT the FTUE preview/result modal flow), `Save & continue ▸` (primary, keeps the existing FTUE setup workflow that lands the user on the connected-models success modal). Edit mode keeps the legacy chrome unchanged. - Mobile responsiveness for the new chrome. The connect form's breadcrumb row and the tappable `Choose provider` step both drop on phone-sized viewports — the AppBar back-arrow already covers the same affordance, so re-rendering the breadcrumb chip on top of the form would just crowd the layout. The header card and step indicator stay; the footer's three buttons reflow onto multiple rows via `Wrap` instead of overflowing. - API-key "Get a key at " hint on the connect form is now a real link target — taps wire through `url_launcher` to open the provider's console in the system browser (`https://` is prepended for the bare-host URLs stored in `aiProviderKeyConsoleUrl`). The hint is also marked as `Semantics(link: true)` and underlined so screen readers announce it as a link, not a plain caption. - Connect-form breadcrumb root crumb now reads `Settings` (using the canonical `settingsV2DetailRootCrumb` key, same as the Settings V2 top-crumbs widget) instead of the unrelated `Theming` label that previously surfaced. Romanian CTAs (`Save and continue`, `Save as draft`, `Saved as draft`) were retoned to the formal `Salvați` register to stay consistent with the rest of the Romanian AI Settings flow, and the base-URL hint replaces the stilted `punctul final oficial` with `punctul final implicit`. ## [0.9.998] ### Fixed - Sync stack hot-path cleanup driven by the 2026-05-10/11/12 desktop slow- and super-slow-query logs. Eight load-bearing offenders folded into one pass: - `BackfillRequestService` no longer runs the 2-minute periodic body against an empty sequence log. A new `hasActionableEntries()` probe (matches the `idx_sync_sequence_log_actionable_status_…` partial indices via `status IN (1, 2) LIMIT 1`) short-circuits both retire passes and `_loadNextUnqueuedMissingBatch` when no `missing`/`requested` rows exist — the 347 no-op ticks/day that each ran ~5 sync_db queries are gone. - `SyncDatabase.getPendingBackfillEntries` switched from drift's parameterised `status.isIn(...)` to a literal `status IN (0, 3)` via `CustomExpression`. The planner can now seek the outbox index instead of falling back to `SCAN outbox` (357 hits/day, avg 226 ms, max 1.8 s on 2026-05-12). - `InboundEventQueue._oldestActiveOriginTs` inlines its status set as a literal `IN ('enqueued','leased','retrying')` so the partial index `idx_inbound_event_queue_active_room_ts` finally matches; the parameterised form had degraded into a rowid scan at up to 862 ms per commit/abandon. - New `AgentRepository.getEntitiesByIds` collapses the per-id `Future.wait(getEntity)` fan-out used by both `ProjectAgentWorkflow._collectObservationPayloads` and `TaskAgentWorkflow._collectObservationPayloads` into a single `WHERE id IN (?, …)` query. `TaskAgentWorkflow._buildLinkedTasks ContextJson` switched to the existing bulk `getLinksToMultiple` + `getLatestReportsByAgentIds` pair — eliminates the compounding N×M `agent_links WHERE to_id = ? AND type = ?` + per-link `getLatestReport` fan-out (2 484 + 2 203 slow hits on 2026-05-10). - `QueuePipelineCoordinator` debounces `AttachmentIndex.path Recorded` events into a 100 ms accumulator and flushes through a new bulk `InboundEventQueue.resurrectByPaths(paths)`. A burst of attachment downloads now resolves in one SELECT + UPDATE round-trip instead of N independent writer-lock transactions (222 super-slow hits/day on 2026-05-12). - `sync_db` schema bumped to v21 with two literal-status partial indices for outbox claim ordering: `idx_outbox_pending_created_id (created_at, id) WHERE status = 0` and `idx_outbox_sending_expiry (updated_at, created_at, id) WHERE status = 3`. The pending-claim path no longer has to pay `USE TEMP B-TREE FOR ORDER BY` after seeking `idx_outbox_status_priority_created_at`. - Agent database schema bumped to v8 with `idx_wake_run_log_created_at`. `getWakeRunsInWindow` had been falling back to a base-table scan + temp B-tree because every existing wake_run_log index was leaded by `agent_id` / `template_id` / `status`. - `MatrixSyncMetricsPanel`'s on-screen poll widened from 2 s to 5 s (now exposed as the `pollInterval` constant). The 30 polls/min cadence had pulled the `GROUP BY status, producer` aggregate to 223 hits/day; widening removes the polling pressure without losing perceived liveness. - Tasks Filter sticky footer no longer renders the Clear all / Save pills as floating text labels on dark mode. The button's inner box was 44 tall while its slot enforced a 56 minimum, so under the glass-blur footer the painted pill drifted out of alignment with the centered text and at low alpha against the dark sheet the silhouette washed out entirely. Both heights now share a single `DesignSystemFilterMetrics.actionMinHeight` token and the non-highlighted variant carries a 1 px border so the pill stays visible against the gradient overlay; Apply keeps its solid accent. ### Added - AI Settings page redesigned to match the new design-system reference. The collapsing v1 title strip now sits above an `AiSettingsHeaderBar` that carries an inline search field; the page's "Add" affordance rides on a per-tab `DesignSystemFloatingActionButton` (Scaffold's `floatingActionButton` slot) instead of a header CTA — see the FAB bullet later in this section for the icon/handler swap rules. The tab row uses an `AiSettingsTabBar` that bakes the counters directly into each label — "Providers 2 / Models 5 / Profiles 2" — instead of a separate counter strip. Below the tab bar the active tab renders the new card surfaces: `AiProviderCard` is a 2-column responsive grid card with the provider's accented icon top-left, a `⋯` overflow menu top-right, the display name + tagline, a hairline divider, and a status row with a colored dot + left label + right meta — the three status variants are Connected (with model count tail), generic "Invalid key · Fix" (covers missing / wrong / revoked / 401 / 403 in one phrase per the design tweak), and "Offline · Make sure Ollama is running"; `AiModelCard` is a single-column row with the provider icon, model name + monospaced provider-model id inline, and the capability chips below (intentionally no on/off toggle); and `AiProfileCard` is a 2-column grid card with the active badge inline with the name and a task→model mapping list (Thinking / Image recognition / Transcription / Image generation → model). Unresolved model ids render as `missing` in the warning tone so dangling profile references are visible at a glance. With zero providers configured the page swaps in `AiSettingsFtueBanner` — an accented banner with a sparkle icon, lead copy, and a "Start setup" button that opens the same provider-type picker — followed by `AiSettingsNoProvidersCard`, a single wrapper card with "No providers yet" + the four first-class provider chips (Gemini / OpenAI / Anthropic / Ollama) that each open the connect form preselected to that provider. Per-provider chrome is wired through a new `AiProviderVisual` helper so the accent / surface / display name / tagline lookups stay consistent across the cards, chips, banner, and (upcoming) provider detail page. ~30 new l10n keys × 6 locales for the page chrome, status copy, capability chips, and the empty / FTUE surfaces. - AI Settings FTUE modals redesigned to match the new design system. The old `FtueSetupDialog` / `FtueResultDialog` are replaced by `AiProviderSetupPreviewModal` and `AiProviderSetupResultModal`. The preview modal lists each model the FTUE will create as a checkbox row the user can untick — those `providerModelId`s are passed to `runFtueSetupForType` as `excludedProviderModelIds` and the per-provider setup helpers skip them at creation time, so the success-modal model count reflects exactly what landed in the database. Models the provider already owns appear in a read-only "Already added" section so re-running the wizard doesn't pretend they're new. Ollama short-circuits the preview because it has no canonical model preset, and providers whose entire preset is already configured skip straight to the result modal. The result modal shows a compact post-setup summary — "{Provider} is connected. We set things up for you." — with bulleted rows for added models, created profile, and test category (created vs reused), an errors block when present, and Review setup / Start using AI buttons. Both modals use the existing app dark / light tokens instead of the design's near-black, and pick up the per-provider accent from the `colors.aiProvider.*` tokens added in the next bullet. - AI Settings FTUE backend now covers Anthropic and Ollama alongside the existing Gemini / OpenAI / Mistral / Alibaba flows. Anthropic setup seeds Claude Sonnet 4 (reasoning) and Claude Haiku 3.5 (fast) plus a new `Anthropic Claude` default inference profile bound to a fresh `Test Category Anthropic Enabled` test category. Ollama setup installs the test category bound to the existing `Local (Ollama)` profile and intentionally creates no model rows — Ollama serves whatever the user has pulled locally, so the upcoming connect modal will enumerate `/api/tags` and let the user pick from what's actually installed. Both providers are now members of `ftueSupportedProviderTypes`, dispatch through `runFtueSetupForType`, and render via `AiProviderSetupPreviewModal` / `AiProviderSetupResultModal`, so the FTUE flow fires end-to-end after a key paste (Anthropic) or a localhost ping (Ollama). The per-provider result types now share a sealed `AiFtueResult` base class so the analyzer catches a missing arm when the next provider is wired in. - Design-system color tokens for AI providers: `colors.aiProvider.{gemini, openAi, anthropic, ollama}` each carry an accent + tinted-surface pair in both light and dark variants, generated from `tokens.json` through the existing token generator. Used by the upcoming redesigned AI Settings surfaces (quick-add provider tiles, provider cards, master/detail rail) so per-provider accents stay consistent across mobile, desktop, and modal layers. - AI Settings provider detail page. Tapping a provider row on the AI Settings page now opens `AiProviderDetailPage` (under `lib/features/ai/ui/settings/provider/`) instead of dropping the user straight into the edit form. The detail page shows a provider-tinted header strip with the display name, tagline, and derived status pill; a Connection card with masked API key (last 4 visible), base URL, and display name plus an `Edit` action; a Models section that lists the provider's own models via `AiModelCard` with an `Add model` button; an Active profile section that surfaces the inference profile most strongly tied to this provider's models (default profile first, otherwise the first match) via `AiProfileCard`, hidden when no profile references the provider's models; and a Danger zone that routes through the existing `AiConfigDeleteService` so cascade-delete + confirmation modal + undo snackbar match the rest of the AI settings. The bottom of the page pads by `DesignSystemBottomNavigationBar.occupiedHeight` so the danger-zone card never slips behind the app's bottom nav on mobile. - Provider card "Fix" affordance now auto-focuses the API key field. When a provider's status is `invalidKey`, the `Fix →` link routes through the provider detail page with `focusApiKey: true`, which immediately pushes `InferenceProviderEditPage(focusApiKey: true)`. The edit page wires a `FocusNode` through `AiTextField` and requests focus on the next frame, so the user lands directly on the field that needs editing in one tap. A regular card tap still opens the detail page without auto-pushing so users who want to inspect models / profile before editing don't get bounced. - Card overflow menus do the right thing. The `⋯` icon on Provider, Model, and Profile cards used to render a disabled IconButton on top of a tappable card — tapping the icon just forwarded the tap to the card itself. Each card now accepts a `menuActions` list (`AiCardMenuAction` / `AiCardActionMenuButton` under `widgets/v2/ai_card_action_menu.dart`) and the AI Settings page passes Edit + Delete rows; Delete runs the existing `AiConfigDeleteService` so cascade-delete + confirmation + undo snackbar are consistent with the rest of the settings. Cards without a populated `menuActions` list (e.g. the cards embedded in the provider detail page) hide the icon entirely instead of showing a non-interactive affordance. - Inference model edit page redesigned to match the v1–v3 visual language. The old SliverAppBar + `colorScheme.surface` chrome is replaced by a clean AppBar with a text `Save` action (the `FormBottomBar` is gone — Cmd+S still saves), a provider-tinted header strip that reads the model name + owning provider inline, and two design-system sections (`Identity`, `Capabilities`) rendered as `level02` cards on a `level01` page background. Provider, input modalities, and output modalities use a shared `_SelectorField` that styles a read-only `AiTextField` as a tap-to-open dropdown and tints the trailing caret amber when the field is empty so unset required selections are visible at a glance. All form copy (section names, field labels, hints, toggle descriptions, "Select a provider" hint, back tooltip, save button) is now localised across all six supported locales — previously the page shipped hard-coded English on every label. - Card text on the Models and Profiles tabs no longer truncates on mobile. The model card now stacks the display name, the monospaced `provider/model-id` line, and the capability chips vertically — previously the name and id sat inline on one row and both ellipsised at narrow widths. The profile card lets the profile name, description, and each task→model slot wrap to multiple lines instead of clipping; slot rows are top-aligned so the icon + slot label stay flush with the first line of a wrapped model name. The inference profile form's body padding now includes `DesignSystemBottomNavigationBar.occupiedHeight` so the last form section clears the bottom nav on mobile. - AI Settings desktop master/detail dispatch. The `ai` panel in the Settings V2 registry was promoted from a plain `AiSettingsBody` body to a multi-kind `AiPanelDispatch` so the right pane on desktop now swaps to `AiProviderDetailPage`, `InferenceModelEditPage`, or `InferenceProfileDetailPage` in place when a row is tapped (or when the user beams to `/settings/ai/provider/`, `/settings/ai/model/`, or `/settings/ai/profile/`) — previously the detail page was pushed as a fullscreen route over the master/detail shell, hiding the sidebar. The three typed URL patterns (`/settings/ai/provider/:providerId`, `/settings/ai/model/:modelId`, `/settings/ai/profile/:profileId`) handle both surfaces: on desktop the panel slot reacts to `NavService.desktopSelectedSettingsRoute`, on mobile the Beamer location builder pushes the detail page on top of the AI Settings page so the back gesture still returns to the list. `AiSettingsNavigationService.navigateToConfigEdit` now beams to the matching typed route instead of calling `Navigator.push` directly, and the Fix-flow rides the provider URL with a `?focusApiKey=true` query parameter — `AiPanelDispatch` reads that query off the live route and rebuilds the detail page with `focusApiKey: true` so the edit form auto-pushes with the API key field focused regardless of which surface the user came in on. - AI Settings desktop sidebar shows Providers / Models / Profiles as their own leaves under "AI Settings". Previously the three views lived only as tabs inside the AI Settings page body, and the sidebar carried a single (legacy) "Inference Profiles" leaf that pointed at the V1 profile list — the v3 redesign moved profile rendering into the tab body but the legacy leaf was never updated. The settings tree now mirrors the three tabs as three leaves, each panel renders `AiSettingsBody` pinned to its tab with the in-pane TabBar hidden, and the `ai-profiles` panel is repointed from `InferenceProfilesBody` (V1) to the v3 Profiles tab body so a desktop user sees one consistent profile list whether they reach it via the sidebar or the AI Settings page directly. Clicking the AI Settings parent row itself also renders without the TabBar — the page-level default tab is already Providers, so the parent landing lands visually identical to the Providers leaf instead of briefly flashing a duplicate tab strip before the user picks a leaf. On mobile the page still ships its TabBar — the sidebar doesn't exist there, so the tabs remain the only way to flip between the three views. - AI Settings "Add" affordance is a per-tab floating action button. The inline "Add provider" pill in the page header is gone — it was hard-coded to provider creation even when the user was looking at Models or Profiles, and it duplicated the FAB pattern the rest of the settings surface already uses (`InferenceProfilePage`, habits, measurables). The page now drives a `DesignSystemFloatingActionButton` in the Scaffold's `floatingActionButton` slot (wrapped in `DesignSystemBottomNavigationFabPadding` so it clears the bottom nav on mobile). Its `semanticLabel`, `icon`, AND handler swap per active tab — `Icons.bolt_rounded` on Providers, `Icons.psychology_alt_rounded` on Models, `Icons.tune_rounded` on Profiles. The icon set mirrors the sidebar leaf icons so the FAB glyph echoes the leaf the user is on and names what's about to be added. `AiSettingsHeaderBar` is now just the search bar — the subtitle paragraph the v2 redesign shipped was redundant with the page title + sidebar leaf, and the custom design-system text input it used had a different placeholder than every other search field in the app. The header now reuses `AiSettingsSearchBar` (which wraps the app-wide `LottiSearchBar`) so the styling and hint copy match the rest of the settings surface. - Provider card's status row right-aligns the model count tail flush against the card's right edge. The v3 layout shipped a `Spacer + Flexible(Text)` pairing for the right-side tail, which split the remaining width 1:1 and parked "3 models · last used 2m ago" in the middle-right of the card instead of at the edge. `Expanded(Text(textAlign: end))` collapses the gap so the tail snaps to the right margin; the Ollama-hint variant on the offline status branch got the same treatment. - Models tab regains the pre-v3 filter strip in full. The v3 redesign hooked up the filter service to filter by `selectedProviders` / `selectedCapabilities` / `reasoningFilter` but never shipped the UI for any of them. The strip now reuses the existing `AiSettingsFilterChips` widget (which itself wraps `ProviderFilterChipsRow(useStyledChips: true)`), so the Models tab gets the same chrome the rest of the app already uses: provider-tinted pills (colored leading dot + matching border + dark accent-tinted background pulled from the per-provider `colors.aiProvider.*` tokens), capability chips (Text / Vision / Audio) plus a Reasoning toggle in a neutral Material FilterChip style, and an appearing "Clear filters" action when at least one filter is active. The chips sit in a `Wrap`, so the row stacks across multiple lines on narrow viewports instead of horizontally scrolling — the design-system-chip prototype shipped earlier on this branch (and its successor with `useStyledChips: false`) didn't match what the rest of the app does and were both removed. - Sync conflict resolution screen rebuilt as an inline-diff picker. Tapping a conflict row now opens a dedicated page with a back chip + title + amber count pill in the header, lead copy, and an amber summary banner that calls out the entity type and how long ago the two sides diverged plus a subline listing the fields that differ. Below the banner are two diff cards — local (teal) on the left, remote (blue) on the right — with each side's title rendered as a word-level diff: tokens unique to the local side are tinted green, tokens the remote dropped are line-through red, and tokens the remote introduced are tinted blue. Each card carries the side's timestamp + per-side vector-clock counter (`vec N`) plus a category icon, word count, and audio duration when available, with a `local edit` / `via sync` provenance label on desktop. A picker pill row underneath mirrors the selection (`Use this device` / `Use from sync`, plus `Edit & merge…` on desktop), and a sticky glass footer holds Cancel + Apply with helper copy that reads back the chosen consequence in plain English. Below 768 px the cards stack vertically, the picker drops to two pills, and the footer surfaces `Edit & merge…` as a left-side text link. Apply commits via `PersistenceLogic.updateJournalEntity` only when a side is picked; Cancel beams back without writing. The diff itself is a pure-Dart word-level LCS helper so it carries unit tests for identical / additions-only / replacements / mixed cases. - New design-system color tokens for the picker: `colors.conflict.*` carry the local / remote / diverged accent + tinted surface pairs, and `colors.diff.*` cover the added / removed / replaced highlight pairs. Both are wired through `tokens.json` and the generator so light and dark variants are emitted from one source. ### Fixed - The Task action bar's inset stop circle now persists the running timer entry's `dateTo` to `DateTime.now()` before the timer service is cleared, instead of only clearing in-memory state. Previously `_onStopTimer` called `TimeService.stop()` directly, which left whatever `dateTo` was last written on disk — typically minutes behind the actual stop tap — so the recorded session was short by the elapsed-but-unsaved tail. The handler now routes through `EntryController.save(stopRecording: true)` for the running timer entry, mirroring the entry-editor stop button. - Slow-query log on the 2026-05-10 desktop super-slow trace flagged four hot paths that all stemmed from query/index mismatches: - `InboundQueue.stats()` ran four separate `selectOnly` aggregates (three of them full SCANs of `inbound_event_queue` with a TEMP B-TREE for GROUP BY) on every `_emitDepth` tick — captured at 1014 ms and 2244 ms in the slow log because no index covered `(status, producer)`. Collapsed to one `GROUP BY status, producer` query and added a v20 `idx_inbound_event_queue_status_producer_enqueued` composite so the pivot is index-only over a tight key range. - `getProjectTaskRollups` (`database.dart`) fell back to `idx_journal_browse + USE TEMP B-TREE FOR GROUP BY` (327 ms) because the query omitted `task = 1` even though every Task write sets it. Adding the redundant predicate lets the planner match the v40 `idx_journal_project_task_status` partial index and stream the aggregate directly. - `ChecklistRepository.getChecklistItemsForTask` materialised every `ChecklistItem` the device had ever seen, JSON-decoded each one and filtered in Dart (558 ms on the agent hot path). Replaced with two indexed bulk-by-id lookups via `journalEntitiesByIdsUnorderedAllPrivate`: first the parent Checklists (whose `data.linkedChecklistItems` already lists their children), then the items themselves. The unused `deletedOnly` parameter was dropped because no caller passes `true`. - `SyncSequenceLogService` cache cleared every cached host watermark, last-seen timestamp and materialised upper bound the moment a single global 5-minute timer ticked over, regardless of which host had been active. That produced the 200–500 ms `getLastCounterForHost` waves visible in the slow log: a quiet host's watermark got wiped just because some unrelated host had been queried 5 minutes earlier. Replaced with a per-host expiry map and a separate global TTL for the entry-keyed last-sent-counter LRU; `_advanceLastCounterCache` now refreshes the host's window so a host being actively backfilled does not expire mid-run. - The Projects tab "+" FAB now opens the project create page directly inside the Projects tab via `/projects/create` instead of beaming through `/settings/projects/create`. On desktop the old route landed on the Settings V2 root (no `projects` panel is registered there), so the create flow was unreachable. The same path swap is applied to the inline "New project" button on category pages so deep-links with a prefilled `categoryId=` continue to work. The legacy `/settings/projects/create` pattern is removed. - The Categories list "+" FAB on Definitions → Categories now opens `CategoryCreateModal` directly instead of beaming to `/settings/categories/create`, which on desktop was bouncing the user back to Settings V2's root. The modal stays inside the current tab and the list refreshes via the shared category stream once the new entry is persisted. - The "Save" / "Create" row on the project create page is no longer hidden behind the bottom navigation pill on mobile. The page's `FormBottomBar` is wrapped in a `Padding` sized to `DesignSystemBottomNavigationBar.occupiedHeight`, so it docks above the pill on phones and remains flush at the bottom on desktop where the helper returns 0. - Project create no longer publishes a "Saved successfully" toast on the projects tab. The freshly-created project shows up in the list immediately on pop (the list watches `projectsOverviewProvider`), which is sufficient confirmation for a one-shot create flow — and avoids the disruptive Material FAB lift that a floating SnackBar would otherwise trigger on the projects list. Error toasts on the create page itself are unchanged and still dock above the form's bottom action bar. - Conflict detail page now opens correctly on desktop. The Settings V2 panel for `sync-conflicts` was registered as a list-only body, so a row tap updated the URL but the right-hand pane kept rendering the list — only the mobile Beamer stack was wiring the detail. The panel now uses `DetailIdDispatch(idParamKey: 'conflictId')` so the desktop pane swaps to `ConflictDetailRoute` when the URL gains an id, matching the categories / labels / dashboards / measurables flows. - Linked-entries activity log on a task now sorts entries by each entry's `dateFrom` instead of the link's own `createdAt`, so the "Newest first" / "Oldest first" toggle reflects the timestamps shown in the row headers. Previously two entries linked in the same session could appear in an order that contradicted their visible times because their links shared a `createdAt` and the list was sorted by that field. The widget now reads a new `sortedLinkedEntriesProvider` that resolves each linked entity and falls back to `link.createdAt` only while an entry is still loading. ### Added - The project create form now exposes a category picker between the title and target-date fields. The picker reuses the shared `CategoryField` and is seeded from the route's `?categoryId=` query parameter so the inline "New project" button on a category page still pre-selects the right category, while users opening the form from the Projects tab FAB can pick a category up front instead of having to backfill it after creation. ### Changed - Sync Conflicts list page restyled to match the agents listing visual language. The legacy card chrome is gone; each row is now a hover-aware inline strip that pairs a tone-driven status badge (success for resolved, danger for unresolved) with an entity-type badge, the creation timestamp as the row title, and an 8-character mono prefix of the conflict id on the trailing edge with the full id revealed via tooltip. The full vector clock — previously wrapping two lines of monospace text inside every row — is no longer surfaced in the list and stays available on the conflict detail page where the merge view actually needs it. Below 600 px the row collapses to a two-line stacked layout so phone-width viewports stay legible. The shared `monoMetaStyle` helper that used to live inside the agents feature was also lifted into the design system so any feature that needs an Inconsolata mono cell pulls from a single source. - AI summary card header (the `AI summary / Task Laura` row) keeps its wake / refresh affordances and Read more pill inline alongside the title on every viewport instead of stacking them underneath on cards narrower than ~360 px. The title block uses an `Expanded` so it softWraps to a second line if the controls really need the room; on a phone-sized card with the wake-cycle countdown active the pill also switches to a tighter compact width so the inline cluster reads less crowded. Previously the header dropped the entire control cluster to a second row even when only the refresh icon + Read more pill were present and would have fit alongside the title comfortably. - Toast notifications triggered from the task details page (e.g. "Change applied" after confirming a proposed change) now float above the sticky `TaskActionBar` on every platform instead of being pinned to the screen / window bottom edge, so the confirmation reads as belonging to the task view it came from. The page wraps its `Scaffold` in a nested `ScaffoldMessenger`, scoping `context.showToast()` calls fired from inside the subtree to that messenger — Flutter then floats the `SnackBar` above the `bottomNavigationBar` slot automatically. Previously this was gated to macOS / Linux / Windows; on mobile the `TaskActionBar` would cover the toast docked at the screen bottom. ## [0.9.997] ### Changed - Sidebar running-timer card now hides itself only when the timer's parent task is the same task currently open in the desktop task- details pane *and* the user is actually on a `/tasks/` route. The detail page's sticky action bar already shows a running indicator, so duplicating the title in the sidebar would be noise — but the card still surfaces on every other tab (Habits, Settings, …) because `desktopSelectedTaskId` is sticky across tab switches and the action bar isn't visible there. The show/hide flip runs through a combined `AnimatedSwitcher` + `AnimatedSize` (~220 ms, `Curves.easeInOut`) so the card fades and the surrounding sidebar collapses smoothly instead of popping. The stream is seeded with `TimeService.getCurrent()` so a session already running is rendered on first frame instead of flashing through a hidden state. - Sticky action bar's running "Track time" pill gets a touch more breathing room on the leading edge: the inset stop circle now sits 8 px from the pill border (was 4 px) so it no longer crowds the edge while the elapsed digits tick. ## [0.9.996] ### Changed - Audio recording player restyled to match the green Figma audio card. The play/pause control is now a soft circular pill drawn on the design-system surface token with the high-emphasis text color for the glyph; the previous gradient + drop shadow + animated progress ring are gone. The waveform's played bars use `tokens.colors.interactive.enabled` (the brand teal `#2BA184` in light mode, `#5ED4B7` in dark mode), with a thin teal scrubber line drawn at the playhead and bars widened to 4 px with 3 px spacing for parity with the Figma reference. The fallback progress bar (used while the waveform is still resolving) follows the same tokens, with the unplayed track now reading from `tokens.colors.decorative.level02` instead of an alpha overlay on `onSurfaceVariant`. The play/pause button drops a step in size (40 px compact / 48 px standard) so it no longer dominates the card. - Task details page now has a sticky action bar pinned to the bottom in place of the floating action button. A primary "Track time" pill starts a new timer when idle. While a timer is running on the open task the pill morphs into a live elapsed-time readout with an inset stop circle on the leading edge — tapping the pill body navigates to the running timer entry (matching the desktop sidebar's timer card), and only the inset stop circle stops the timer. The duration uses the same tabular-figures / slashed-zero / cv02–04 font features as the sidebar timer; under one hour it's compacted to `mm:ss` and only expands to `hh:mm:ss` once a session crosses the hour mark. Four round affordances cover the remaining frequent inline actions: record audio (immediately after the pill — turns red with a white glyph while a recording for the open task is active), add checklist, import image, and more actions (opens the previous menu for long-tail items, including capture-screenshot on desktop platforms that support it). Capture-screenshot is no longer a top-level affordance. On narrow viewports the row stays on a single line by dropping affordances in priority order: image first, then checklist; both remain reachable via the more menu. On mobile the bar takes over the bottom edge entirely: the app shell's bottom navigation pill is hidden whenever the active beamer route is `/tasks/`, letting the action bar dock flush against the home indicator. ## [0.9.995] ### Changed - Agent UI/UX refinements across the AI summary card, sidebar Wake Queue, and the renamed "Wake Cycles" settings page. The AI card's outer glow is trimmed to a soft tinted edge instead of a wash, the underlined subtitle below "AI summary" now reads the agent template name (e.g. "Task Laura") rather than the generic kind label, the Confirm-all button pins to the right edge of the proposals header to line up with the Read more pill above it, and the cards inside the agent internals panel (Reports / Conversations / Observations / Activity) now sit on the AI panel's row token so the dark backdrop reads as a unified surface. Settings → "Pending Wakes" is renamed to "Wake Cycles", its countdown is now zero-padded MM:SS (or HH:MM:SS once over an hour) with tabular figures so digits no longer breathe, refresh keeps the previously-loaded rows on screen instead of blanking to a spinner, and a new "Running now" block at the top of the page surfaces the currently-executing wake instances with a live elapsed pill. The desktop sidebar Wake Queue gains an Ongoing block (live duration since wake start), only shows scheduled wakes due within the next hour (anything farther out is hidden from the sidebar entirely), drops the per-row letter avatar, uses smaller mono font with ellipsis for long titles, prefers the linked task / project title as the row label, and the WAKES header itself is now an open-in-new link straight to the full Wake Cycles page. ## [0.9.994] ### Changed - Desktop running-timer is now an inline sidebar section in the `aboveSettings` slot, replacing the bottom-anchored floating indicator. The new panel shows the running task title (up to two lines, in the design-system caption style), a ticking HH:MM:SS counter, and a stop button. Tapping the body navigates to the running task (or the timer's journal entry); tapping the stop button stops the timer. The mobile bottom-nav overlay is unchanged. The timer text uses Inter with tabular figures, slashed zero, and the `cv02`/`cv03`/`cv04` open-digit variants so 4/6/9 stay legible and digits don't breathe. - Sidebar nav-item vertical spacing tightened from `step6` (24 px) to `step5` (16 px) — a 33 % reduction — to make room for the new timer section without the rail feeling crowded. ### Fixed - Task detail "Linked from" no longer surfaces parent projects. Tasks link to a project as their organising context, but listing the project alongside other linked entries adds noise without the user ever wanting to act on it; `LinkedFromEntriesWidget` now filters out `ProjectEntry` items unconditionally. ## [0.9.993] ### Changed - Task filter modal action bar now uses the new "Apply filter" glass footer: a full-width frosted-glass strip flush to the bottom of the modal, with a hairline divider on top and right-aligned Clear all, Save, and Apply filter buttons. The last filter section gains comfortable breathing room above the footer; long localized labels still fit because the buttons can grow past their default width. ### Fixed - Task filter modal's Save button now applies the current modal draft, persists it to the saved-filters sidebar, and closes the modal — one action instead of three. Previously Save read from `liveTasksFilterProvider` (the last applied state), so any edits made in the modal before tapping Save were silently dropped from the persisted filter, and the modal stayed open afterwards. The design-system filter modal's `onSavePressed` callback now receives the committed name AND the current draft state, so any consumer can derive a payload from what's actually visible to the user. ### Changed - Saved-filter "Saved / Updated / Deleted" confirmation toasts now render via the shared design-system toast (`context.showToast`) — same rounded pill, palette, and queue behavior the label-detail page uses when a label is created — instead of an ad-hoc themed [SnackBar]. The leading glyph is the standard tone-coloured `check_circle_rounded` from `DesignSystemToast`. ### Changed - Re-skinned Settings → Agents → Pending Wakes onto the shared listing shell. The legacy stacked `_PendingWakeCard` column is replaced by the same toolbar / row primitives Instances, Templates, and Souls use: search, an optional Type filter (only when both `pending` and `scheduled` records exist), Group by All / Type, and Sort by Due soonest (default) / Due latest / Name. Each row shows the linked task or project title (or the agent name when no subject is attached), agent kind + wake type pills, and a live countdown chip that ticks once a second (the absolute due timestamp surfaces as the chip's tooltip). Internal: the per-row `Timer.periodic` is replaced by a single page-scoped `wakeCountdownTickerProvider`; rows derive their visible countdown via `ref.watch(... .select(...))` so 1K rows collapse to one timer and only rebuild when their displayed string changes. New `agentPendingWakeRowVmsProvider` joins `pendingWakeRecordsProvider` with `pendingWakeTargetTitleProvider` (one parallel `Future.wait`) into `PendingWakeVm`s the page maps into the shared `AgentListRowData`. Removed the legacy `agent_pending_wakes_list.dart` and the inline `WakeActivityChart` from this tab — the chart still lives under the Stats tab as a 24h overview. - Re-skinned Settings → Agents → Souls onto the shared listing shell. The legacy ModernBaseCard / ListTile column is replaced by the same toolbar / row / search primitives Instances and Templates now use: search input, Sort by Name (default) / Recent / Oldest, and a hue- tinted `SoulAvatar` initial-tile leading. Each row shows the soul name and the active version as a mono `vN` cell. Tap target (deep- link to the soul detail) and the create-soul FAB are unchanged. Internal: new `agentSoulRowVmsProvider` joins `allSoulDocumentsProvider` + `activeSoulVersionProvider` into `SoulVm`s the page maps into the shared `AgentListRowData`; the old inline `_SoulsTab` / `_SoulListTile` widgets are removed. ## [0.9.992] ### Added - New inline `SidebarWakeQueue` block on the desktop sidebar's `aboveSettings` slot (handoff S1), gated by the new `show_sidebar_wake_queue` config flag (default off). Renders a `WAKES N` header, the next two pending wakes as compact `avatar · agent · ETA` rows, and an `Open list` / `+N more` link that opens the full Pending Wakes view. ETAs are localized: `now`, `mm:ss` under an hour, `Xh MMm` beyond — and switch to the warning colour inside the last five minutes. Each row has a trailing `×` button that calls `cancelPendingWake` (pending) or `clearScheduledWake` (scheduled). Tapping the row body deep-links into the agent's instance page. When the queue is empty the rows collapse but the header stays visible so the section never feels stuck. ### Changed - `SyncActivityIndicator` is now a single fixed-width row (`• tx N · • rx N`) pinned to the bottom of the sidebar via the new `belowSettings` slot, so the inline Wake Queue can take its previous position above Settings. Each numeric column reserves a fixed slot — values right-aligned inside it — so the LEDs and labels never reflow as the outbox / inbox depths roll over between digit counts. - Re-skinned Settings → Agents → Templates onto the shared listing shell. The legacy ListTile + ModernBaseCard list is replaced by the same toolbar / group / row primitives the Instances tab now uses: search input, optional Kind filter (only when 2+ kinds exist), Group by Kind / All, Sort by Name / Recent / Oldest. Each row shows the template name, model id, kind pill, and active version (`v{N}`) in the mono meta cell; a pending-review template gets an `AgentPalette.purple`-tinted leading icon instead of the legacy positioned dot. Tap target and FAB behaviour are unchanged. Internal: new `agentTemplateRowVmsProvider` joins `agentTemplatesProvider` + `activeTemplateVersionProvider` + `templatesPendingReviewProvider` into `TemplateVm`s the page maps into the shared `AgentListRowData`. - Extracted the Settings → Agents → Instances scaffolding (page shell, toolbar, group section, row, filter chip row, soul avatar) into a shared `lib/features/agents/ui/listing/` layer so the Templates, Souls, and Pending Wakes tabs can be re-skinned through the same primitives in the next pass. Public API: `AgentListingShell`, `AgentListRowData` (with a closed `AgentListPillTone` enum and a `Widget Function(BuildContext)?` `trailing` slot), `AgentListLeading` (sealed: avatar / icon), `AgentListFilterAxis` / `GroupAxis` / `SortAxis`, and the pure `buildGroupedAgentList` pipeline. `AgentInstancesPage` is now a thin adapter that maps `InstanceVm` into the shared row VM; rendering, filter / sort / group state, and the empty / loading / error branches all live in the shell. This scaffolding extraction is internal-only — the Instances redesign below is what users will see. - Redesigned Settings → Agents → Instances. Replaced the filter strips + card list with a denser layout: a toolbar with multi-select Filters (Type / Status / Soul), Group by (Soul / Type / Status), Sort (Recent / Oldest / Name), live search and a result counter; active filters surface below as removable chips with Clear all. Each row shows name + template + ID + type and status pills + last-activity time, under a collapsible group header that surfaces per-group active and total counts. The page now uses background level-02 to match the rest of the settings surface; the AppBar reads "Agent Instances" on this tab. - Responsive layout: at narrow widths the toolbar wraps to multiple lines and the search input takes a full line below; instance rows stack title / template / ID, then pills + time + chevron, on two rows. - Search field focus is painted on the outer border (teal accent + glow) instead of an inner Material underline. - Internal: `_typeLabel` / `_statusLabel` helpers, `SoulOption`, and `hueForSeed` consolidated into one place; per-page filter counts bundled into a cached `FilterCounts` so they're computed only when the row list changes. ## [0.9.991] ### Added - New unified `AiSummaryCard` on the task details page replacing the separate AI summary panel and decision activity strip. Single deep-teal-tinted-navy surface that hosts the agent TLDR, an inline expandable Goal / Achieved / What's left / Learnings report under a Read more / Show less pill, the actionable proposals list with swipe + button confirm-or-reject and a Confirm-all batch, a collapsible `History · N` toggle for resolved proposals, and a recent-activity footer (See activity / Hide activity) capped at six rows. The wake-cycle affordances (running spinner, run-now refresh, countdown pill, cancel-timer) sit directly in the header. - New `AgentInternalsPanel` right-side overlay reachable from the agent name link, the footer avatar, or the "Open agent internals" pill. Clamped to 600–800 px on wide screens; on narrow screens (phones in portrait, slim split-view windows) it renders as a full-screen modal that slides up from the bottom. Hosts the same five tabs (Stats / Reports / Conversations / Observations / Activity) as the standalone agent detail page via the new shared `AgentInternalsBody` widget. - Design-system tokens for the AI surface: `color.aiCard.*` (card background, raised surface, row, borders, accent + accent-soft, footer washes, body / meta / faint-meta text) and `color.proposalKind.{add, update, remove, priority, estimate, status, label, due}.{color, surface}`. The token generator now emits these new groups and the matching `DsColorsAiCard` / `DsColorsProposalKind*` classes. Five new localized strings cover the activity row's relative time labels (now / minutes / hours / days / weeks). ### Changed - `task_form.dart` now mounts `AiSummaryCard` instead of the old `AgentSuggestionsPanel`. The legacy `AgentSuggestionsPanel` and `TaskAgentReportSection` widgets and their tests are removed (no remaining call sites). - `AgentDetailPage` body extracted into the new shared `AgentInternalsBody` so the standalone page and the side panel render the exact same tabs. - Agent and tasks feature READMEs updated to describe the new card, the side panel, and the shared body. The tasks README's "Visual surface" section now distinguishes standard `TaskDetailSectionCard` cards from the dedicated AI surface and links to a new follow-up doc, `docs/design/missing_density_typography_tokens.md`, which catalogs every place the card overrides `height` / `letterSpacing` on top of base typography tokens and proposes a `Compact/*` density tier. - Settings root list is shorter and friendlier for new users. Habits, Categories, Labels, Dashboards, and Measurables now sit under a single new "Definitions" entry; Config Flags moved into the Advanced sub-page next to Logging, Maintenance, and About. Existing deep links (`/settings/habits`, `/settings/flags`, …) are unchanged so bookmarks keep resolving. ### Fixed - Checklist body no longer leaves a fat empty band between the Open / Done / All filter strip and the first item on devices with a top safe-area inset (notched iPhones). The plain `ListView.builder` introduced in #3060 was inheriting the ambient `MediaQuery.padding.top` as its own top padding; it now passes `padding: EdgeInsets.zero` explicitly so the first row sits flush under the filter divider on every form factor. - Linux geolocation rewritten as a dual-mode `package:dbus` client. Inside the Flathub sandbox `DeviceLocation` calls `org.freedesktop.portal.Location` (`XdgLocationPortal` in `lib/services/linux_location_portal.dart`); the portal mediates GeoClue on the app's behalf and Lotti now appears under GNOME Settings → Location → Permitted Apps. Outside the sandbox (`flutter run`, dev builds) it talks to `org.freedesktop.GeoClue2` on the system bus directly (`LinuxGeoClueClient` in `lib/services/linux_geoclue_client.dart`), because the portal rejects unsandboxed callers with `Access denied` (the portal keys authorization off the caller's Flatpak/Snap app-id, which is empty for host processes). The picker keys off `/.flatpak-info`. The previous code path used the abandoned `geoclue` Dart package, required `` to be replaced by the caller, never appeared in GNOME's permitted-apps list, and had no `--talk-name` entry in the Flatpak manifest. The `geoclue` dependency is removed and the manifest now declares `--talk-name=org.freedesktop.portal.Desktop`. Note: GeoClue itself still depends on a working WiFi-based geolocation backend. On Ubuntu 24.04 the default Mozilla Location Service URL returns 404 since Mozilla deprecated MLS in 2024; point GeoClue at BeaconDB by dropping into `/etc/geoclue/conf.d/00-beacondb.conf`: ```ini [wifi] enable=true url=https://api.beacondb.net/v1/geolocate submission-url=https://api.beacondb.net/v2/geosubmit ``` then `sudo systemctl restart geoclue`. Without a working backend Lotti silently falls back to IP geolocation on every platform that uses GeoClue under the hood. ## [0.9.990] ### Added - Activity-filter pill row above the linked entries section on the task details page. Three toggleable pills (Timer / Audio / Images) hide or show the corresponding linked entry types; checklists keep their dedicated section at the top of the page. Pills follow the Figma spec — accent fill at 15 % alpha, accent border, accent label when active; design-system tokens drive Timer; Audio and Images use the Figma hex values until those colors land in the token set. - Sort + filter modal opened from a "Newest first / Oldest first" trigger on the right of the activity bar. Reuses the `DesignSystemFilterChoicePill` and `DesignSystemFilterPalette` from the task list filter modal so both surfaces share one chrome. Includes a "Show hidden entries" switch using the existing include-hidden controller. ### Changed - Linked entry cards now use `TaskDetailSectionCard` (the existing design-system card already used by the linked tasks section) instead of `ModernBaseCard`. Card padding aligned with the task list card — `tokens.spacing.step4` (12 px) horizontal, `step4` vertical at the bottom and `step1` (2 px) at the top so the trailing IconButtons in the entry header provide the visual breathing room without doubling it. - Audio player no longer wraps itself in a second card. The redundant `ModernBaseCard` was removed so the player sits flush inside the entry's section card with just a thin vertical pad. - Checklist card header padding tightened (`step1` top; collapsed state also drops bottom to `step1`) and the chevron is now a proper `IconButton` with `Collapse` / `Expand` tooltips, so it gains a hover state matching the adjacent menu button. - "Checklists" subheading removed from the task details page. The per-task sort menu still appears (right-aligned) when there are two or more checklists. - Extracted `DesignSystemFilterChoicePill` from the private task filter sheet into the shared `design_system_filter_shared.dart` so both the task filter modal and the new linked-entries filter modal render the same exclusive-choice pill — fill cross-fade, accent border alpha tween, constant border width, 400 ms animation. - Linked tasks section on the task detail page is redesigned to match the Figma spec: a single expandable card with a count badge in the header, per-row "to"/"from" direction glyphs (info-blue / success-green), status circles, and dividers between rows. The overflow menu retains the existing link/create/manage actions. ## [0.9.989] ### Added - AI assistant icon and skill menu now appear on text journal entries. Skills that consume entry text (coding-prompt generation, image-prompt generation, cover-art generation) work the same way whether the source is a typed note or a voice recording's transcript. - New built-in **Generate Design Prompt** skill. Turns task context plus the entry's notes into a UI/UX design exploration prompt that defaults to five functional prototypes (override-able from the entry text), enforces design-system alignment when one is mentioned, and surfaces clarifying questions up front. Output is two-section Markdown ready to paste into Claude / Figma Make / v0.dev. - New built-in **Generate Research Prompt** skill. Produces a structured Markdown research brief (Background, Research Questions, Scope, Deliverables, Source Preferences, expected output format, open questions) ready to paste into Claude with Research or ChatGPT Pro with Deep Research. ### Changed - Built-in skills are now defined in code under `lib/features/ai/skills/built_in_skills.dart` and read via `skillRegistryProvider` instead of being seeded into the AI config DB. This removes the seeding round-trip and lets skill content stay in lockstep with the code that uses it. A future skill-management UI will introduce a separate user-override layer rather than re-introducing seeding. - Generated-prompt cards now render the source skill's own name (e.g. "Generate Design Prompt", "Generate Research Prompt") instead of always showing "AI Coding Prompt". `AiResponseData` now carries an optional `skillId` so the same card can distinguish sibling skills that share the `promptGeneration` response type. ## [0.9.988] ### Changed - Outbox post-drain settle window raised from 250 ms to 1500 ms and lifted into `SyncTuning.outboxPostDrainSettle`. Bursty edits (rapid typing, imports, multi-entity flows) now coalesce into the next bundle instead of shipping one bundle per write, so the outbox sends fewer, fuller trains while keeping the first-departure latency unchanged. ### Fixed - Faster task-list, projects, and inbound sync queries via new SQLite indices and tighter sync transactions. - New partial index `idx_journal_tasks_status_priority_date` streams the open-tasks list ordered by `(task_priority_rank, date_from DESC)` without falling back to a temporary B-tree, even when many categories are selected. - New covering `idx_linked_entries_from_id_hidden_to_id` resolves the bulk linked-time-spans query index-only on the linked side. - New `idx_inbound_event_queue_status_enqueued` lets the queue-stats `COUNT(*) + MIN(enqueued_at)` poll seek by status and pick MIN from the index instead of scanning the full queue ledger. - `claimNextOutboxBatch` no longer scans the outbox: the original `status = pending OR (status = sending AND updated_at < cutoff)` predicate is now two indexed seeks merged in Dart, each picked up by the existing actionable partial index. - Sync apply for journal entities holds the journal writer lock only around the actual writes. Cross-database awaits (sync-sequence log recording, pre-write `journalEntityById` reads) now run outside the journal transaction so concurrent readers no longer queue for hundreds of milliseconds per applied event. - Removed the per-open self-heal block that re-issued `CREATE INDEX IF NOT EXISTS` on every database connection. With drift's read-pool the block ran nine times per launch (writer plus eight read isolates) recovering from an incident that has not occurred in production. `beforeOpen` is now just `PRAGMA foreign_keys = ON`; if a migration is ever interrupted on a future device the next schema bump will repair it. ## [0.9.987] ### Fixed - Inbox backfill no longer hangs for ten minutes per agent-bundle row. Removed the agent wake-cycle bundling layer that coalesced per-wake agent writes into a `SyncAgentBundle`. With generic dequeue-time outbox bundling in place, the wake-bundle envelope was redundant — and after the outbox bundler started claiming text-only rows in batches it became actively harmful: bundled `SyncAgentBundle` rows shipped a manifest whose envelope-only entries no longer pointed at any uploaded attachment, leaving inbound rows stuck in the 10-minute pending-attachment wait until they were skipped. Agent writes now hit the outbox directly as `SyncAgentEntity`/`SyncAgentLink` rows and flow through the same bundling path as journal entities. The `SyncAgentBundle` wire variant remains parseable so messages from peers that predate this change still decode; the receiver no-ops them and any children missing locally resurface via the existing per-(host, counter) backfill path. ## [0.9.986] ### Fixed - Reduced idle CPU usage after starting and stopping task timers. Inactive desktop and mobile navigation tabs now disable tickers for their cached `IndexedStack` children, and task/project agent countdowns update only their small countdown pill instead of rebuilding whole report sections in hidden tabs. Running-timer task icons also rebuild only when the linked task changes. - Inbound sync rows waiting for missing attachment or agent-bundle JSON now use a bounded 10-minute wait with long backoff instead of either retrying forever or being skipped by the generic attempt counter too early. If the attachment is still missing after the grace window, the row is abandoned and normal backfill recovery can handle the long tail. ## [0.9.985] ### Added - Sidebar **Sync activity indicator** (variant D4a). When the new `show_sync_activity_indicator` config flag is enabled, the desktop sidebar gains an ambient two-row monospace strip just above Settings: `tx ` / `rx `, each with a 5×5 LED that flashes for ~140 ms per packet committed on that channel (TX = events uploaded to the homeserver, RX = inbound events applied locally). Tap navigates to Settings → Sync. When the flag is on, the legacy red `289` Settings badge is suppressed so the count is shown in exactly one place. Hidden in the collapsed sidebar — the strip is too narrow there. Default off; opt in via Settings → Config Flags. ## [0.9.984] ### Added - Task agents can now propose user-reviewed edits to historical time entries linked from the current task. The wake prompt exposes editable non-running entry IDs with their current time range and text, the new `update_time_entry` tool validates task ownership and excludes the live timer, and the suggestion row renders a current-to-proposed diff before confirmation. ### Changed - `create_time_entry` completed sessions now allow any valid `endTime > startTime` range, including future and midnight-spanning blocks. Running timers keep the existing today-only and not-in-the-future restrictions. - DailyOS due-task hot path (`getTasksDueOn` / `getTasksDueOnOrBefore`) now reads a denormalized `due_at INTEGER` column on `journal` instead of `json_extract(serialized,'$.data.due')`. The partial `idx_journal_tasks_due_open` is rebuilt on the column so the planner streams `ORDER BY due_at ASC` straight from the index without per-row JSON parsing, eliminating the planner-fragility that required the previous `INDEXED BY` pin plus JSON-fallback safety net. The `due_at` shadow is populated by `toDbEntity` on every upsert and back-filled for every existing task with a non-null `data.due` — including completed and rejected tasks — by the v41 migration. The serialized JSON payload is unchanged; the column is purely a query-acceleration shadow. - DailyOS calendar prefetch (`sortedCalendarEntries`) now coalesces concurrent per-day callers into a single union-range DB round-trip with client-side filtering, modeled on the existing `_coalesceOpenTasksDueUpTo` pattern. Slow-query telemetry showed 22+ identical date-range queries firing in the same second during multi-day prefetch; the wave collapses them into one read. - `retireExhaustedRequestedEntries` and `retireAgedOutRequestedEntries` now flip rows in `pageSize`-bounded batches, each in its own transaction, so a large backlog never holds the sync-DB writer lock through a single multi-second UPDATE. Previously a stuck-row backlog could pin the lock for ~1.9 s, starving concurrent reads. ### Fixed - Improved sync reliability between devices. ## [0.9.982] ### Added - New `DsPill` design-system primitive (`lib/features/design_system/components/chips/ds_pill.dart`) with `filled`, `tinted`, `outline`, and `muted` variants plus `DsGhostChip` and `DsDividerDot` companions. Shared anatomy: 28 px tall, pill radius (`radii.badgesPills`), `spacing.step3` horizontal padding, `spacing.step2` leading/trailing gap, `body.bodySmall`-sized label with optional override, optional leading/trailing slots, and Material ink-well that supports both `onTap` and `onLongPress`. The muted variant paints a 1 px dashed border via a private `CustomPainter` rather than pulling in `dotted_border` as a dependency. - New `CreateChecklistItem` on the FAB's create-entry menu (`create_entry_action_modal.dart`). When the FAB sits on a task detail page (i.e. `linkedFromId` resolves to a `Task`), the menu now surfaces *Add Checklist* as the first entry; on every other surface the item self-hides. ### Changed - **Task detail header — Option B redesign** (`docs/design/design_handoff_task_header/`, `lib/features/tasks/ui/header/desktop_task_header.dart`, `lib/features/tasks/ui/header/desktop_task_header_connector.dart`). The classification + metadata Wrap with pipe separators is replaced by a two-tier hierarchy: - A breadcrumb above the title — `▣ Category / Project name` — gets the categorical "where am I?" info out of the chip soup. The 10×10 rounded category square is the *only* place the category color is used as a fill. Project name truncates with ellipsis; "No project" placeholder remains tappable. Each segment has a `surface.hover` background, no pill chrome. - A single horizontal pill row carries the *actionable* metadata (priority, due, estimate, labels). Priority is `DsPill.tinted` in its accent (P0 = error, P1 = warning, P2 = info, P3 = success) with the production SVG glyph at 14 px. Due is `DsPill.outline` that flips to `DsPill.muted` "No due date" when null and keeps the urgency tinting (overdue → error, today → warning, otherwise `text.mediumEmphasis`). Labels render as filled pills with an 8 px color dot; long-press still surfaces the description dialog. The `+ Add Label` ghost chip is shown only when no labels exist — once labels are present, tapping any one opens the same selector. - The status select is rewritten as `_StatusPill` with a per-status 18 % alpha tint (in-progress → info, blocked → error, on-hold → warning, groomed → interactive accent, done → success), neutral tint for "open", and a strikethrough low-emphasis treatment for "rejected". - `_TrailingAlignedWrap` — a custom multi-child render box — replaces the old `LayoutBuilder`-with-Row/Column branch. Leading chips wrap greedily; the status pill is pinned to the right edge of whichever row it lands on, falling onto its own right-aligned line only when it doesn't fit on the last leading row. No breakpoint-driven layout switch. - All chips read `tokens.typography.styles.others.caption` (12 px) so priority through status share one rhythm. The estimate label uses `text.lowEmphasis` to defer to the running tracker. - The estimate chip in the connector is rebuilt on top of `DsPill`: `DsPill.muted` "No estimate" when unset, `DsPill.filled` with a 36 × 6 progress bar in the trailing slot when running, `DsPill.tinted` in the error color when overtime. - **Tasks list filter chips** typography unified with the detail header — `ActiveFilterChip` now uses `others.caption` (12 px) and the trailing remove icon is sized to 20 px. - **Task list cards** — when the user has narrowed the filter to a single status, the trailing status pill is now omitted on every card (`tasks_tab_page.dart` passes `showStatus: selectedTaskStatuses.length != 1` to `TaskBrowseListItem`). With zero or 2+ statuses selected the chip stays as a disambiguator. - **Checklists section on the task detail page** — `lib/features/tasks/ui/checklists/checklists_widget.dart`: - The inline `+` "Add Checklist" button on the section header is removed. Adding the first checklist now lives exclusively on the FAB's create-entry menu. - When the task has zero checklists, the entire section (header + sort menu) collapses to `SizedBox.shrink` so the empty state doesn't dangle. - The populated section now sits with a `tokens.spacing.step5` (16 px) gap above it so it doesn't crowd the previous block. ## [0.9.981] ### Added - Outbox message bundling, gated by the new `useOutboxBundlingFlag` config flag (default off). When enabled, `OutboxProcessor` claims a contiguous run of consecutive text-only outbox rows up to `SyncTuning.outboxBundleMaxSize` (50, tuneable up to 100) in `(priority, createdAt)` order and ships them as a single `SyncMessage.outboxBundle` envelope delivered as a sidecar JSON attachment under `/outbox_bundles/.json`. Media-attachment rows always travel alone — the batch claim stops one before the next attachment row. Receivers download the sidecar via the existing attachment-ingestor pipeline and unpack it through `OutboxBundleUnpacker`, dispatching each child through the per-type apply path used for individually-delivered messages. Bursty workloads (agent suggestions, checklist creation, sync backfill replay) drop to roughly 1 Matrix event per ~50 rows. See ADR 0015 for the design, rollout, and per-child fault-isolation details. ## [0.9.980] ### Added - Categories settings list now matches the Labels / Dashboards baseline. A `DesignSystemSearch` field above the list filters rows by name (case-insensitive substring); a "+" floating action button at the bottom-right opens the existing `/settings/categories/create` route. The legacy in-header "Add Category" text button is gone — the FAB is the canonical create affordance now and uses the design-system circular FAB component already consumed by the Tasks tab. New arb keys `settingsCategoriesSearchHint`, `settingsCategoriesNoMatchQuery`, and `settingsCategoriesCreateTitle` cover all six locales; the no-match empty state echoes the active query verbatim (`No categories match ""`) so users know what they typed. - `DesignSystemToast` grows two new affordances and is now the shared toast surface for the migrated call sites: - `ToastAction(label, onPressed, semanticsLabel?)` renders an inline call-to-action (e.g. "Undo") in the trailing slot. Action and dismiss can coexist; the action label inherits the tone's border color so the CTA reads as part of the tone, not a generic link. - `countdownDuration` paints a thin tone-coloured progress strip along the toast's top edge that drains over the supplied duration; `initialCountdownProgress` lets a resumed countdown pick up mid-bar. - `BuildContext.showToast` learns matching `action`, `countdown`, `initialCountdownProgress`, `replaceCurrent`, and `clearQueue` parameters; a new parallel extension `ScaffoldMessengerState.showDesignSystemToast` exposes the same surface to async callbacks that may outlive the calling widget (e.g. swipe-to-dismiss `onDismissed`, where the row is removed from the tree before the toast fires — captured `messenger` survives the unmount, captured `context` does not). `replaceCurrent: true` only hides the *current* SnackBar (items already queued behind it still appear afterwards); `clearQueue: true` calls `clearSnackBars()` so terminal-status toasts (e.g. "all confirmed") supersede every queued per-item toast. - The shared SnackBar wrapper auto-extends its visible duration by one second when `countdown: true` so the bar reaches zero before the SnackBar fades. - Migrated `ScaffoldMessenger.showSnackBar` call sites that were semantically tone-coded notifications now go through `DesignSystemToast`: - Checklist item swipe-to-delete + swipe-to-archive (warning tone, 5s / 2s countdown, Undo action) — replaces the old `showCountdownSnackBar` helper. - Checklist correction-capture pending toast (success tone, countdown matches `kCorrectionSaveDelay`, Cancel action) — the `CorrectionUndoSnackbarContent` widget now renders `DesignSystemToast` internally and ticks every 500 ms only to refresh the "save in N s" title; the countdown bar runs off a single animation controller so parent rebuilds don't stutter it. - Agents: per-row Suggestion confirm / reject and AgentSuggestions panel confirm-all (success / warning / error tones based on result), with strings + messenger captured before the await so the toast still fires after the row is unmounted by the suggestion list rebuild. Confirm-all and per-row result toasts use `clearQueue: true` so terminal status replaces any in-flight per-item toasts. - Sync: Backfill Sync `Catch up now` and `Retry skipped events`, plus the QueueDepthCard `Retry all skipped` action (success on completion, error on throw). Out of scope and not migrated in this release: the `AiConfigDeleteService` cascade-deletion SnackBar (custom rich body) and the saved-task-filter pill toast (custom rounded chrome). ### Changed - Settings sidebar order: Sync now sits directly below Agents in the Settings V2 tree (root order: `whats-new`, `ai`, `agents`, `sync`, `habits`, `categories`, `labels`, `dashboards`, `measurables`, `theming`, `flags`, `advanced`). Agents and Sync are both runtime / system concerns and read better as a pair than separated by the taxonomy leaves (habits / categories / labels). - Settings V2 desktop now exposes the **provisioned-sync** (QR pairing) entry point. The Sync branch was previously leafless on V2 desktop, so the `ProvisionedSyncSettingsCard` that the mobile `SyncSettingsPage` already renders had no equivalent surface — provisioned setup was effectively unreachable on desktop. The Sync branch now declares `panel: 'sync'` and a new `_syncPanel` builder in the registry hosts the same card inside a `SyncFeatureGate` + `DesignSystemGroupedList` so the visual treatment matches the rest of the V2 detail pane. - Settings list-page FABs now use the design-system circular FAB component (`DesignSystemFloatingActionButton`) — the same visual primitive the Tasks tab already consumes — instead of raw Material `FloatingActionButton`. Migrated sites: `LabelsListPage`, `DefinitionsListPage` (the shared host for Dashboards / Habits / Measurables, via the public `FloatingAddIcon` helper), `AgentSettingsPage` (Templates + Souls tab FABs), `InferenceProfilePage`, and `AiSettingsFloatingActionButton`. The previously-extended AI FAB drops its inline per-tab text label in favour of the circular shape; the per-tab label survives as the `semanticLabel` so screen readers and hover tooltips still announce `Add Provider` / `Add Model` / `Add Profile` correctly. (Categories already adopted the DS FAB in the parity commit above.) ### Removed - `lib/widgets/misc/countdown_snackbar_content.dart` (and its test) — the `CountdownSnackBarContent` widget and its `showCountdownSnackBar` helper had no callers left after the checklist swipe flows moved to `DesignSystemToast`'s native `countdown` + `action` API. ## [0.9.979] ### Added - Keyword search on the Config Flags page. A `DesignSystemSearch` field above the list filters flags by their localized title or description (case-insensitive, leading/trailing whitespace trimmed). When no flag matches, a centered empty-state message replaces the list. The filter logic lives in a pure `filterDisplayedFlags` function so it can be unit-tested independently of `BuildContext`. The list itself is wrapped in design-system spacing tokens (`step2` vertical, `step4` between the field and the list) so the search affordance reads as part of the same surface rather than a floating chrome. - The `enable_whats_new` config flag is now a user-toggleable entry in the Config Flags list (icon, title, and description sourced from the new `configFlagEnableWhatsNew{,Description}` arb keys across all locales). The DB-side flag definition was already in place — this surfaces the toggle in the UI so users can hide the What's New tree leaf without editing the database. ### Changed - Config Flags page polish — four follow-ups on top of the keyword search: - The `DesignSystemSearch` field is now pinned at the top of the page while only the list of flag rows scrolls beneath it. `FlagsPage` opts its `SliverBoxAdapterPage` chrome into a new `fillRemaining: true` mode that hosts the body inside `SliverFillRemaining(hasScrollBody: true)` and folds the bottom-nav-occupied space into the body padding. `FlagsBody` itself becomes `Column[fixed search, Expanded(scrollable list)]` so the same restructure works inside the V2 detail pane (`'flags'` panel registry entry flips to `scrollable: false` since the body now owns its scrolling). - `DesignSystemListItem` gains a `subtitleMaxLines` parameter (default `1`, preserving the existing single-line ellipsis for every other caller). Flag rows pass `null` so long descriptions like "Generate AI summary for task actions" wrap onto a second / third line instead of truncating. Honored on both the plain-text `subtitle` and the `subtitleSpans` paths. - The flag list now hides the divider on hover — `_FlagsList` becomes stateful, tracks `_hoveredIndex` via the existing `DesignSystemListItem.onHoverChanged`, and suppresses the divider both *below the hovered row* and *above the next row* so the hovered row is never bisected by a hairline. Mirrors the polish already used in the task-list rows. - Added a horizontal page gutter and consistent vertical spacing so the search field and the rounded list container both honour the same edges; cards no longer sit flush against the screen edge. - Backfill Sync settings page rebuilt around the Option C "compact ledger" design: a welded status row (Inbound queue · Missing · Skipped) at the top, a leader-dotted Sync statistics ledger, an `Automatic backfill` toggle, and a single collapsed `Advanced recovery` group containing every manual recovery action (`Catch up now`, `Retry skipped events` when present, `Manual backfill`, `Reset unresolvable`, `Re-request pending`, `Ask peers for unresolvable`, `Retire stuck entries`). All visuals are driven by design-system tokens — `DesignSystemButton`, `DesignSystemToggle`, `tokens.colors.*`, `tokens.spacing.*`, `tokens.radii.*` — so the page reads as part of the rest of Settings V2 instead of a one-off. - Settings V2 (tree-nav + detail pane) is now the canonical desktop chrome. `SettingsRootPage` reduces to a plain mobile→`SettingsPage` / desktop→`SettingsV2Page` dispatch with no flag observation; mobile is unchanged. - Settings V2 Agents tab is now URL-driven and the sidebar tree mirrors the in-page tab strip. Each tab gets a dedicated tree leaf under `Agents` in the same order as the in-page TabBar — `agents/stats`, `agents/templates`, `agents/instances`, `agents/souls`, `agents/pending-wakes` — and the `templates` / `instances` / `souls` panels are wrapped in `DetailIdDispatch` so the "+" FAB and row taps swap the right-hand panel between list / detail / create. `SettingsLocation` registers the bare `/settings/agents/{stats,templates,instances,souls,pending-wakes}` patterns so Beamer accepts them without falling through to a parent location, and `SettingsTreeUrlSync` carries a symmetric URL-driven guard alongside the existing tree-driven one so URL → tree sync preserves panel-local trailing segments (`/create`, a detail UUID) instead of canonicalizing them away. The in-page tab strip (`_AgentSettingsTabBar`) is hidden when `NavService.isDesktopMode` is true — exposing both navigation surfaces caused the URL → tree → URL feedback guard to leak across rapid tab clicks (silently swallowing subsequent sidebar / FAB / row beams) and let the sidebar selection drift out of sync because a top-tab click never expanded the parent branch in the sidebar. The body still resolves its content from the URL, so dropping the bar on desktop is purely additive; mobile / push- stack callers keep the legacy local-`setState` behavior. - `DetailIdDispatch` forces `StackFit.expand` on its inner `AnimatedSwitcher` so Scaffold-based panel bodies (`AgentSettingsBody`, …) receive bounded constraints and lay out their `floatingActionButton` the same way they would unwrapped. ### Removed - The `enable_settings_tree` config flag and its description, the legacy desktop `SettingsColumnStack` view, the `DisableV2Button` escape hatch (and its embeds in `EmptyRoot` / `DefaultPanel`), and the matching arb keys (`configFlagEnableSettingsTree{,Description}`, `settingsV2DisableAction`, `settingsV2DisableFailed`) across all locales. The `placeholderButtonIconSize` constant is gone with its sole caller. ### Fixed - Category selection modal now closes itself when a row is tapped on mobile and narrow desktop. The bottom-nav redesign opens these modals on the **root** Navigator (above the bottom nav), but the picker callbacks were popping `Navigator.of()`, which resolves to the per-tab nested Navigator the page lives in rather than the root Navigator that hosts the modal. Selecting a category therefore left the picker stuck open while the underlying per-tab stack popped. The four affected callers (`DesktopTaskHeaderConnector._showCategoryPicker`, `CategorySelectionIconButton`, `CategoryField`, `AddBlockSheet`) now name their `builder` argument `modalContext` and pop with `Navigator.of(modalContext).pop()` — the same pattern commit c6627fe8d already applied to the estimate / due-date / priority pickers, and that `ProjectDetailsPage._pickCategory` was already using. ## [0.9.978] ### Changed - Settings V2 detail surface consolidated to a single page-title surface. The full breadcrumb path (e.g. `Settings › Sync › Backfill Sync`) now lives in the page header at Heading 3 typography — non-leaf segments are tappable and call `SettingsTreePath.truncateTo(depth)`, mirroring the click rules of the old in-pane crumbs. The new `SettingsV2TopCrumbs` reads ids from `settingsTreePathProvider` and resolves them against the shared `SettingsTreeScope.index`, falling back to just the root label when the scope isn't mounted. `LeafPanel` now returns the cached `IndexedStack` directly: the in-pane crumb trail, the Heading-3 leaf title, and the outer `step6 / step5` `Padding` gutter are all gone, so registered panels fill the detail pane edge-to-edge. The `_LocalCrumbs` and `_CrumbLink` helpers were deleted along with the `LeafPanel` test cases that exercised them; the remaining tests assert the IndexedStack matches the `LeafPanel` size (no outer Padding) and that no chevron renders inside the leaf subtree. - Sync conflict resolution moved out of the **Advanced** branch and into the **Sync** branch in the Settings V2 tree. The leaf id is now `sync/conflicts` and the panel registry key is `sync-conflicts`. The Beamer URL stays at `/settings/advanced/conflicts` (and the `:conflictId` / `:conflictId/edit` subroutes are unchanged) so existing deep links and the legacy column-stack layout keep working — only the V2 tree shape and the panel registry / label resolver follow the reorg. The Sync branch is now always visible regardless of `enableMatrix`: the matrix-only leaves (backfill / stats / outbox / matrix-maintenance) are still flag-gated, but Conflicts stays reachable so legacy or local-only conflicts can always be resolved. The localized subtitle key was renamed `settingsAdvancedConflictsSubtitle` → `settingsSyncConflictsSubtitle` to match the new branch. - Task agents now react immediately to recorded audio. Brand-new task agents on a blank task no longer surface a "wake in 2:00" countdown — the orchestrator skips the throttle deadline while the agent is awaiting content, and the content gate keeps the agent from running until real content arrives. The moment a transcription finishes (cloud or realtime), the task agent is nudged via a manual wake that bypasses the standard throttle, so the user does not wait through a 2-minute window after speaking. Subsequent typing edits still go through the normal 2-minute coalescing throttle. Implemented via a new `WakeReason.transcriptionComplete` and an `awaitingContent` mirror in `WakeOrchestrator`, populated by `TaskAgentService` on creation and restoration. ## [0.9.977] - 2026-04-25 ### Review polish (Settings V2) - `SettingsTreeView` and `SettingsDetailPane` now share a single tree + `SettingsTreeIndex` via the new `SettingsTreeScope` inherited widget hosted by `SettingsV2Page`. The five config-flag subscriptions and the `buildSettingsTree` call happen exactly once per page mount instead of being duplicated in each consumer, and the two views can no longer disagree on the tree shape. - Leaf panel bodies are now kept mounted across sibling-leaf switches: `SettingsDetailPane`'s `AnimatedSwitcher` uses a stable `'leaf'` key and `LeafPanel` caches visited bodies behind an `IndexedStack`, so scroll position, filter state, and in-flight async loaders survive when the user pings between sibling entries (e.g. `agents/templates` ↔ `agents/souls`). - Panel registry switched from an ad-hoc `_scrollable(...)` wrapper to a declarative `SettingsPanelSpec(build: ..., scrollable: ...)` so each body opts in to the outer `SingleChildScrollView` at the registration site — contributors can't accidentally double-scroll a `CustomScrollView`-bearing body any more. - `SettingsTreeUrlSync` replaces its single `_programmaticBeam` bool with a counter so overlapping rapid taps don't leave the URL → tree guard stuck or released early between post-frame callbacks. - `SettingsTreeNavWidth` splits its `_userAdjusted` flag into a user-intent flag and a persisted-load-race flag, guards against state writes after dispose, and cancels queued persists from the debounce timer when the notifier has been disposed. - `SettingsTreeIndex.build` now reports duplicate node ids via a pluggable `duplicateReporter` callback (defaults to `debugPrint`), so duplicates remain visible in release builds and not just in debug asserts. `ancestors(id)` returns the pre-wrapped unmodifiable view stored on the index rather than allocating a fresh copy per read. - `beamUrlToPath` canonicalizes URLs by stripping `?query` and `#fragment` before prefix matching, so settings URLs carrying query parameters resolve to the correct leaf instead of collapsing to the empty root. - Misc cleanup: `SettingsTreePath` uses stdlib `listEquals`, the escape-hatch fallback now writes the shared `enableSettingsTreeFlagDescription` constant rather than an empty string, and the breadcrumb link uses design-token spacing and radii in place of hardcoded 4/2 dp values. ### Fixed - Bottom navigation bar polish: the sync outbox badge on the Settings tab now sits in the top-right of the icon to match the Tasks badge, and the audio and time recording indicators dock flush against the visible nav-bar pill instead of floating above it. - Desktop linked-task navigation no longer takes over the full window. Tapping a linked task from a task's details now pushes onto a per-pane stack inside the right-hand details area, leaving the task list pane visible. The back arrow at the top of the task details only appears on desktop when a linked task is on top of the stack, and clicking it pops back to the previous task instead of being a no-op. Mobile navigation is unchanged. ### Added - Saved filters for the Tasks tab. Build a filter (status, priority, category, agent, etc.) in the Tasks Filter modal, tap the new **Save** button next to Apply, give it a name, and pin it under "Tasks" in the expanded sidebar. Saved filters live in `SavedTaskFiltersController` (Riverpod, `keepAlive: true`), persist as a JSON list under `SAVED_TASK_FILTERS` in `SettingsDb`, and surface as a treeview with active-state highlight, hover-revealed two-tap delete, double-click inline rename, drag-to-reorder, and a "Tasks · {filter name}" indicator in the tasks pane header. Activating a saved filter applies it via `JournalPageController.applyBatchFilterUpdate` while preserving the ephemeral search query. ## [0.9.976] - 2026-04-25 ### Changed - Agent wake cycles now sync as one bundled Matrix payload instead of emitting one outbox row per agent entity/link mutation. `AgentSyncService` wraps wake execution in a zone-local `AgentWakeSyncInterceptor` keyed by the wake run, buffers only agent entity/link messages, deduplicates by id, preserves superseded child vector clocks via `coveredVectorClocks`, and flushes a descriptor-backed `SyncAgentBundle` to the outbox when the wake exits. Receivers resolve the bundle attachment, apply entities before links through the existing agent sync handlers, and record each child payload in the sequence log so the convergence/backfill contract stays unchanged while wake traffic collapses to one Matrix message. - `runInWakeCycle` is reentrant-safe (nested calls reuse the active interceptor) and routes nested-transaction post-commit messages into the wake buffer so multi-transaction wakes still ship as a single bundle. - The success-path bundle flush now swallows and logs outbox failures via `DomainLogger` instead of failing the wake run — the database writes have already committed, and maintenance/backfill paths can resurface the committed rows. The failure-path flush continues to attempt one final bundle send before re-throwing the original wake error. - The interceptor's merge step only carries forward the SUPERSEDED clock (`previous.vectorClock`), since `OutboxService._prepareAgent{Entity,Link}` already adds the current clock to `coveredVectorClocks` downstream and the receiver's `_filterCoveredVectorClocks` strips it again before pre-marking covered counters. - Matrix verification emoji sequence now wraps onto multiple lines on narrow phone viewports. The 7-emoji row was overflowing horizontally on Samsung devices in portrait (`RIGHT OVERFLOWED BY 24 PIXELS`), cutting off the last cell. `VerificationEmojisRow` now uses a `Wrap` layout so the cells flow onto a second line when the screen width is constrained while keeping centered alignment on wide screens. - Resync Settings (Sync → Maintenance → Re-sync messages) now has an "Entity types" section with two checkboxes — "Journal entities" and "Agent entities" — both on by default. Unticking "Agent entities" lets the user bypass the agent-entity sweep when catching up a fresh device, where the agent corpus can otherwise enqueue ~20k entries and dominate the backfill cost. The Start button disables and an error hint shows when neither checkbox is selected; `Maintenance.reSyncInterval` enforces the same contract and emits a single `MAINTENANCE/reSyncInterval` log line if the call is invoked with both flags off. - Linux: emojis no longer render as tofu in agent reports, journal text, or any widget consuming a design-system text token. The token generator now emits `fontFamilyFallback: ['Apple Color Emoji', 'Segoe UI Emoji', 'Noto Color Emoji']` on every `TextStyle` so widgets that bypass the `ThemeData`-level fallback (e.g. `AgentMarkdownView`) still resolve the correct emoji glyphs. The global theme's emoji fallback was widened from Linux-only to all non-web platforms — fontconfig (or the equivalent on each OS) ignores missing families so listing all three is harmless. The previously-broken `linux/install_emoji_fonts.sh` now ships its `flatpak/75-noto-color-emoji.conf` companion so the script runs without errors on Ubuntu/Debian dev boxes. - New attachment family `/agent_bundles/.json` is recognized by `isAgentPayloadPath`, the matrix stream helpers' `extractJsonPathFromEvent`, and the queue-apply adapter's transient attachment-error filter. Wake run keys are URL-encoded in bundle file paths to keep them within the documents directory. ## [0.9.975] - 2026-04-25 ### Added - Task Agent now sees the active running timer when one is running. If the timer is for the task being woken, the agent gets full details (timer id, started, live tracked range, elapsed minutes, current entry text) and can propose a richer description via the new `update_running_timer` tool instead of stacking a parallel `create_time_entry` — user-gated, replaces the timer entry text outright on approval, and keeps the in-memory `TimeService` snapshot in sync with the persisted `entryText`, `dateTo`, and `updatedAt`. If the timer is for a different task, only the live tracked range is exposed (no id, no other-task identity, no entry text) so the agent can avoid proposing `create_time_entry` intervals on this task that overlap with what is already being tracked elsewhere. Source-task ownership is checked before the `timerId` comparison so the active timer id is never echoed across task boundaries. ### Changed - `create_time_entry` no longer rejects completed sessions whose start day is earlier than the wake day. The agent can now log a session the user dictates from yesterday or further back; the same-day constraint between `startTime` and `endTime` (no entries spanning midnight) and the not-in-the-future cutoff at wake time are preserved. - `PersistenceLogic.updateJournalEntityText` now returns `false` when its catch block fires, mirroring the contract of `updateJournalEntity`. A caught exception during the update no longer surfaces as a silently-true result with only a logged exception, so callers (including the new `update_running_timer` flow) can rely on the boolean to decide whether to proceed with downstream side effects. ## [0.9.974] - 2026-04-25 ### Changed - Task Details typography pass: the entry editor, AI summary (TLDR + expanded report), Task Agent reports/conversations markdown, linked-task titles, and agent suggestion items now pull font family, sizes, and weights directly from the design-system tokens. The editor body and agent markdown body share `body.bodySmall` (Inter, 14pt, w400); editor headings map to `heading.heading3` / `subtitle.subtitle1` / `subtitle.subtitle2`; linked-task titles use `subtitle.subtitle2`. Code blocks gain a subtle theme-aware surface (`background.level02` lifted by the `surface.enabled` overlay), a 1-px low-emphasis border, and DS radii (`radii.xs` inline / `radii.s` block). The Task Agent markdown checkbox is now compact, non-interactive (no hover overlay or click cursor), and shares the checklist row's `radii.xs` rounded shape, low-emphasis border, and `interactive.enabled` fill. - The AI Coding Prompt / AI Image Prompt card on Task Details now renders both its TLDR summary and its expanded full-prompt body through the shared `AgentMarkdownView`, so its typography, links, and code-block surfaces match the agent report and entry editor (`body.bodySmall` body text, DS heading mapping, low-emphasis checkbox style) instead of the package-default `GptMarkdown` styling. ## [0.9.973] - 2026-04-24 ### Fixed - `VectorClockService` now advances its persisted counter only after the associated write + outbox enqueue both succeed. A new reservation API (`reserveNextVectorClock` returning a `VcReservation`, plus a `withVcScope` zone-based wrapper) bumps the in-memory watermark synchronously so concurrent reservations stay collision-free, but `commit` is what persists to `SettingsDb`; `release` rewinds the in-memory watermark when the reservation was the latest. The known burn call sites — `PersistenceLogic.createDbEntity`, `updateDbEntity`, `updateJournalEntity`, `createLink`, plus `AgentSyncService.upsertEntity`/`upsertLink`/`insertLinkExclusive` and the outermost `runInTransaction` — are now wrapped in `withVcScope` so an `applied=false` rejection, a transaction rollback, or an exception anywhere between `getNextVectorClock()` and `outboxService.enqueueMessage()` rolls the counter back instead of burning a gap that only backfill could close. `getNextVectorClock()` stays available for low-stakes callers — outside a scope it still auto-commits; inside one it auto-attaches. Follow-up to the atomic claim fix in this release, which closed the merge-send race but could not help counters burnt before any outbox row was ever created. - Outbox processor now atomically claims the next row (pending → sending) via `OutboxRepository.claim()` instead of reading it as `pending` and sending it with a stale snapshot. Under the old path, a merge that fired during the ~hundreds of ms of send I/O would `updateOutboxMessage` the row in place (matching `status=pending`), the processor would still send the pre-merge payload and mark the row `sent`, and the merged `coveredVectorClocks` — including the old VC the merge existed to preserve — were silently abandoned. The result was scattered single-counter holes that the receiver flagged as missing and that only backfill could resolve. With the atomic claim the row is `sending` during send, merges' CAS-on-pending update no longer matches, and the merged content is spilled into a fresh pending row that still rides its own Matrix event. The legacy `fetchPending + refreshItem` pair is gone; the actor-side `OutboundQueue` already used the same claim and is now on a shared repository contract. ## [0.9.972] - 2026-04-21 ### Added - Settings V2 foundations behind the `enable_settings_tree` config flag (desktop only; default off). First three steps of the A2 tree-nav + detail-pane rollout documented in `docs/design/settings/settings_v2_implementation_plan.md`: the domain layer (`SettingsNode`, `NodeBadge`, `NodeTone`, `buildSettingsTree(...)`, `SettingsTreeIndex` with `findById` / `ancestors` / `pathToBeamUrl` / `beamUrlToPath`); the state layer (`SettingsTreePath` notifier implementing the four click rules from spec §3 plus `syncFromUrl` / `truncateTo` / `clear`, and a `SettingsTreeNavWidth` notifier that clamps to 280-480 dp with a 300 ms debounced persist under the distinct `SETTINGS_TREE_NAV_WIDTH` key); and the chrome (`SettingsV2Page` with a fixed 56 dp header, a locale-backed `SettingsTreeView` rendering every flag-gated root node via `SettingsTreeNodeWidget`, the `SettingsTreeResizeHandle` with drag / double-tap reset / arrow-key ±8 / shift-arrow ±32 / Home-to-default, and a detail-pane placeholder with a persistent "Disable Settings V2" button — the escape hatch back to the legacy column stack, since no real panels are wired up yet). `SettingsRootPage` picks the surface via a flag watch on `enableSettingsTreeFlag`; with the flag off the legacy multi-column stack is unchanged. - Stray `/settings/maintenance` pathPattern in `lib/beamer/locations/settings_location.dart` replaced with the canonical `/settings/advanced/maintenance` — every caller already routed through the latter, and the A2 tree index now points at the single canonical URL. The old `/settings/maintenance` URL is kept as a pattern alias that renders the same `MaintenancePage`, so any hand-edited bookmark that hit the advertised pattern on `main` keeps working. ### Changed - `BackfillRequestService` now skips analysis+dispatch while the `BridgeCoordinator` is mid-walk (forward-reading fresh timeline events from the last applied event id). Gaps observed during the walk may be closed by events still in the pipe, so the periodic timer and `nudge()` wait for the walk to conclude before asking peers for anything — preventing bogus ~100-entry requests that raced ahead of the inbound path. Exposed via a new `QueuePipelineCoordinator.isBridgeInFlight` getter forwarded from the bridge's existing `_inFlightBridge` state. Manual `processFullBackfill` bypasses the gate so a user action is never silently dropped. - Sync V2 is now the only inbound pipeline. The `use_inbound_event_queue` config flag, its Flags-page toggle, localized strings, and the `suppressLegacyPipeline` switch on `MatrixService` are gone. `QueuePipelineCoordinator` is a required constructor argument, the consumer's live ingestion is always suppressed, catch-up routes unconditionally through the coordinator's bridge, and the legacy connectivity-driven `forceRescan` coalescing plus the `_maybeStartQueuePipeline` flag-branching are removed. Settings seed, CI database expectations, and the integration-test helpers were trimmed to match. - Checklist "Add a new item" pill border now lights up on focus. The previous fix that silenced the themed inner outline left the outer pill visually static on tap, removing the last affordance that the field was actually editable. `_AddItemField`'s outer `AnimatedContainer` now listens to `widget.focusNode` and cross-fades its 1 px border between `tokens.colors.decorative.level01` (idle) and `tokens.colors.interactive.enabled` (focused) over 200 ms easeInOut. Border width stays pinned at 1 px so the pill never breathes across the transition, and sibling rows in the checklist don't shift. - Desktop Settings is now a horizontally scrollable multi-column layout instead of a two-pane list + detail. Every meaningful level of the route tree renders as its own column: `/settings/sync` stacks `SettingsPage + SyncSettingsPage`, `/settings/sync/backfill` adds `BackfillSettingsPage` as a third column on the right, and label/category/dashboard/measurable/habit/agent drill-downs work the same way — up to five columns for `advanced/conflicts//edit`. Columns share a single fixed width (the previously-draggable `listPaneWidth`, now pinned so the rest of the layout stays put), the rightmost column expands to fill any remaining viewport, and the row becomes horizontally scrollable with an auto-scroll to the newly-added column when the stack no longer fits. Drilling from Sync into Backfill adds the third column on the right instead of replacing the Sync sub-menu. Mobile keeps the single-page push navigation fallback. - Settings list rows now match the task list's subdued treatment. `DesignSystemListItem`'s default activated-row fill moved from the saturated `tokens.colors.surface.active` to the new shared `DesignSystemListPalette.activatedFill` — `interactive.enabled` at 12 % opacity, the same source the task-list palette now consumes — so selection reads as a gentle hint rather than a loud block. The component also gained an `onHoverChanged` callback and a `dividerColor` override; the settings root menu uses both to coordinate divider visibility across rows, fading the 1 px line between any two rows to transparent whenever either is hovered or active. The divider keeps its reserved 1 px of vertical space, so hover never causes the column to jitter, and hover tracking is keyed by item identity rather than list index so feature-flag toggles can't strand the suppression on the wrong row. - Queue-pipeline `InboundWorker` now fans the adapter's prepare phase out across the whole batch via `Future.wait`. Prepare is I/O-bound (attachment downloads, gzip decode, JSON decode) with no shared state, so running it in parallel collapses each batch's prepare critical path to the slowest entry instead of the sum. Apply still runs sequentially inside `journalDb.transaction` to preserve the M1 writer-lock discipline, and prepared payloads are cached by `eventId` so apply consumes each one exactly once. Directly addresses the cold-start catch-up throughput cliff where every entry waited on its predecessor's attachment download before its own prepare could begin. ### Changed - Reconnect bridge now attempts a forward-walk from the last applied event id. Previous reconnect behaviour walked the SDK's cached timeline backward from newest-cached to oldest-cached; when the cache already reached months below the current `lastAppliedTs`, every page trivially crossed the boundary on page 0 and the bridge stopped without hitting the server — so a reconnecting client had to rely on peer backfill responses to close gaps in the `[lastAppliedTs, now]` window. The bridge now calls `room.getTimeline(eventContextId: lastAppliedEventId)` (Matrix `/rooms/{id}/context/{eventId}`) for a fresh server-side slice anchored at the marker, then `timeline.requestFuture()` paginates `/messages?dir=f` toward the tip. Backward walk remains as a fallback when no anchor is known or when the server cannot resolve the anchor. Unverified in production — needs observation against a real reconnect with a known gap before we can claim it actually closes the peer-dependent case. - Outbox drain no longer stop-and-go-crawls through large backlogs. `_drainOutbox` was capped at 20 passes per runner callback, after which it bounced back through `ClientRunner.enqueueRequest` — which meant the `UserActivityGate.waitUntilIdle` check and the runner's FIFO queue traversal fired once every 20 items instead of once per drain. `OutboxProcessor` also fetched 10 rows per call via `fetchPending(limit=10)` and threw nine away to send only the oldest. The pass cap is now 2000 (effectively "drain everything in one pass, with a pathological safety net"), and the processor fetches just 2 rows (the head-of-queue plus a cheap `hasMore` probe). Together, a thousand-row outbox that previously took ~50 runner re-entries + 9000 wasted row reads now drains in one runner pass with zero waste — the observable symptom being outbox throughput that looks pathologically slow under bulk enqueues. - Queue pipeline now drops self-echoed events at the live-timeline ingress. Every message this device sends via the outbox comes right back through Matrix's `/sync` on the same room — the legacy pipeline consulted `SentEventRegistry` to short-circuit those echoes, but the new queue pipeline was missing the check, so every outbox send turned into an extra inbound enqueue + prepare + apply + attachment download. Under a large outbox drain this quietly doubled the DB pressure and blocked genuine peer events behind thousands of no-op-self-echoes, which is what made outbox throughput look pathologically slow. `QueuePipelineCoordinator._handleLiveEvent` now consumes from the shared `SentEventRegistry` before touching the queue or the attachment ingestor, and logs a coalesced summary line (one every 30s) with the suppressed count. - Reconnect bridge no longer exits prematurely when the SDK's local timeline cache hadn't loaded the wake-up window yet. Before the fix, a single boundary-crossing page whose events were all duplicates/filtered-out still satisfied the "boundary reached" stop condition, leaving gaps in `[untilTimestamp, now]` to be filled by the slower backfill cadence. `collectHistoryForBootstrap` now keeps paginating past the boundary for up to `boundaryContinuationCap` (5) extra `/messages` round-trips when the sink reports `accepted=0`, giving the SDK a chance to pull more history into its cache. Paired with this, `QueuePipelineCoordinator` now records a "barren bridge" signal whenever a reconnect walk finishes with `boundaryReached` and zero total accepted events — the precise signature of a wedged cache — and the next sequence-log gap detection triggers a one-shot unbounded `collectHistoryForBootstrap` to close the hole aggressively instead of waiting for the normal backfill cadence. Single-flight guarded so a burst of gap signals coalesces onto one walk, and the signal expires after 5 minutes so a stale wedge from hours ago cannot hijack a later gap. - Checklist "Add a new item" pill no longer sprouts a second outline when focused. `_AddItemField` wraps a `TextField` in a container that draws its own 1 px pill border, but only set `border: InputBorder.none` on the inner field — so the app's shared `InputDecorationTheme.focusedBorder` (a 2.5 px primary-colour outline) still overlaid itself inside the pill on tap. Every state-specific border (`enabledBorder`, `focusedBorder`, `disabledBorder`, `errorBorder`, `focusedErrorBorder`) plus `filled` / `fillColor` are now explicitly neutralised on the decoration, so the field stays visually flat inside the pill in all focus states. - New "Ask peers for unresolvable entries" action on the Backfill Settings page. Flips every `unresolvable` sequence-log row back to `missing` so the normal backfill sweep re-asks peers — covers the case where the originating host is dead but a currently-alive peer still has the payload. Complements the narrower "Reset Unresolvable" action, which only resets rows whose `entry_id` is already known locally (a tiny subset after a bulk retirement event). Confirmation dialog with row count before the flip. - New "Retire stuck entries" action on the Backfill Settings page. Runs `retireAgedOutRequestedEntries(amnestyWindow: Duration.zero)` after a confirmation dialog, promoting every currently-open `missing`/`requested` sequence-log row to `unresolvable`. Bypasses the 7-day amnesty window for the case where a user has identified stuck rows blocking the watermark and wants immediate recovery without waiting for the periodic sweep. - `agent_links` upsert no longer fails when a sync-incoming `soul_assignment` or `improver_target` link arrives with a new `id` but the same `from_id` (soul) or `to_id` (improver) as an existing active row. `insertOnConflictUpdate` only handles primary-key conflicts, so the partial unique indexes (`idx_unique_soul_per_template`, `idx_unique_improver_per_template`) threw `SqliteException(2067)` — apply was classified retriable, retries exhausted after 10 attempts, and the queue row was silently abandoned. `upsertLink` now preemptively soft-deletes the colliding active row under the same transaction before inserting, matching the v6 migration's duplicate-cleanup pattern. Soft-deleted rows stay in the table as tombstones for audit. - Stuck `sync_sequence_log` rows no longer block the watermark indefinitely. A row can slip into `requested` via the backfill-response-hint path (which never sets `last_requested_at`) or age out of the active backfill window before hitting the request-count cap — in either case the row sat in a non-terminal status forever, preventing `getLastCounterForHost`'s contiguous prefix from advancing and causing every new event on the same host to re-emit the same gap range through gap detection. A new age-based retire runs alongside the existing exhausted-retire on every backfill sweep: any `missing`/`requested` row older than `SyncTuning.backfillAmnestyWindow` (7 days) is promoted to `unresolvable` regardless of `request_count`. Observed state on a real pair of devices: 3 rows on desktop and 4 on mobile, all created April 8, permanently blocking the watermark — they now retire on the next sweep and gap re-detection stops. - Outbox no longer grows unbounded. `status = sent` rows are pruned after 7 days; `error` rows are kept forever so persistent failures remain inspectable, and `pending`/`sending` rows are never touched. Observed before this change: 395k sent rows on desktop, 265k on mobile — a direct contributor to slow outbox enqueue/dedup queries and heavier WAL checkpointing. The sweep runs at startup (after a 30-second grace so it doesn't contend with init) and every 24 hours thereafter. ## [0.9.969] - 2026-04-21 ### Fixed - Task title edit affordance restored. The new `DesktopTaskHeader` rendered the title as plain read-only text and showed nothing when a task had no title, leaving users with no obvious way to tap into the inline editor. The read-only title now renders a trailing pencil (`Icons.edit_outlined`) next to the label, and empty titles display a localized "No title" placeholder in the same tappable slot. Task list rows in turn render the existing `taskUntitled` "(untitled)" label in the design-system error color + italic so missing titles stand out rather than silently collapsing the row. ## [0.9.968] - 2026-04-20 ### Fixed - Three Phase-2 queue-pipeline gaps flagged in the design review: (1) when the sync room is picked after `MatrixService.init()` or the user switches rooms, `MatrixService.saveRoom` now calls a new `QueuePipelineCoordinator.onRoomChanged` hook that seeds the new room's `queue_markers` row and prunes rows belonging to previous rooms — otherwise the worker replays stale rows against the new room and the new room never gets a marker. (2) On cold start under the queue flag, `QueuePipelineCoordinator.start()` now fires a background `bridge.bridgeNow()` pass once the worker and bridge are attached, mirroring the 300 ms startup `forceRescan` the legacy pipeline runs — reconnects whose timeline is not flagged `limited` no longer silently drop events delivered during login. (3) The F7 `stop(drainFirst: true)` path now uses a new `drainUntilEmpty` loop that flushes the decryption pen, sleeps until each retry lease matures via `InboundQueue.earliestReadyAt`, and re-peeks until the queue is empty or `drainUntilEmptyTimeout` (30 s) elapses — the previous single `drainToCompletion` returned as soon as no rows were ready at call time, stranding retriable / decryption-pending / noRoom rows across a flag-off restart. ### Added - Phase 0 sync diagnostic observability (additive logs only; no behaviour change). `MatrixStreamSignalBinder` now records two probes on the sync room: a `sync.limited` line with `prevBatch`, `eventCount`, and the `sinceMs` gap since the previous sync response whenever the Matrix server returns `timeline.limited == true`; and an `onTimelineEvent.ordering` summary every 100 events tracking strict reorderings and same-timestamp ties. Both route through the `MATRIX_SYNC` log domain into the dated `sync-*.log` file. Data feeds the Inbound Event Queue redesign (`docs/sync/2026-04-20_inbound_event_queue_design.md`). - Stateful inbound sync queue (Phase 1 + 2), gated on the `USE_INBOUND_EVENT_QUEUE` settings flag (default off). When the flag is on, live events, the `limited=true` bridge, and the new "Fetch all history" / "Catch up now" actions route through an `InboundQueue` (Drift-backed, stored in `sync_db`) + `InboundWorker` with a single drain loop per room. Per-room markers live in a new `queue_markers` table so commit + marker advance is atomic and monotonic via `TimelineEventOrdering.isNewer`. Encrypted events are held in a `PendingDecryptionPen` so pre-decryption ciphertext never lands in `raw_json`. The legacy `MatrixStreamSignalBinder` is suppressed when the flag is on, keeping the two pipelines mutually exclusive. Sync Settings gains a flag-gated queue section with a depth card, a Catch-up-now button, and a Fetch-all-history dialog with per-page progress and cancel. Pipeline-tagged log lines (`pipeline=queue` / `pipeline=legacy`) let the Phase-2 validation gate compare apply rates between the two paths. ### Fixed - Desktop UI freeze during Matrix sync. The per-event apply loop in `MatrixStreamProcessor` used to hold the SQLite writer lock while awaiting attachment downloads, gzip decodes, and disk reads — on an 20-event catch-up slice this blocked user-driven journal saves for seconds at a time. `SyncEventProcessor` now splits into `prepare` (all I/O, no DB writes) and `apply` (pure DB writes); the pipeline runs prepare for a chunk in bounded-concurrency batches (4 at a time) outside the transaction, then opens a single transaction to apply the pre-resolved results. Outbox send-side `gzip.encode` is also offloaded to a worker isolate for payloads above 2 KB. - Regression test guards the invariant that every `prepare` completes before the writer transaction opens. - Live sync stalling until force-restart. `MatrixStreamLiveScanController`'s `_scanInFlight` guard could remain set indefinitely if the apply pipeline hung on an unbounded `downloadAndDecryptAttachment()` — every subsequent timeline signal then coalesced into `_liveScanDeferred` and dropped. Two complementary fixes: (1) every Matrix attachment download is now wrapped by `downloadAttachmentWithTimeout` (default 45 s, `SyncTuning.attachmentDownloadTimeout`) which converts a hang into a `FileSystemException` routed through the existing retry tracker; (2) `scheduleLiveScan` is now a stuck-scan watchdog: if a scan has been in-flight longer than `SyncTuning.liveScanStuckThreshold` (90 s), the guard is released, `liveScan.stuck.released` is logged, and a fresh scan is scheduled. Both fixes are independent safety nets; together the pipeline recovers on its own rather than requiring a restart. ### Changed - Sync log volume reduced by ~40% at info level. The outbox send path used to emit ~10 info lines per successful send across four subdomains (`sendMatrixMsg`, `sendNext()`, `outbox.send`, `OUTBOX queue`); a single capture at the outbox processor now summarises the send with `type`, `subject`, `retries`, elapsed `ms`, and remaining `pending` queue depth. Per-event `attachment.observe` and `attachmentIndex.record` lines are now gated behind a `verboseLogging` flag on `AttachmentIngestor` / `AttachmentIndex`; production wiring passes `false` so the steady-state and pagination-burst line counts drop, while tests keep the flag on to preserve existing per-event assertions. Failure, retry, and circuit-breaker diagnostics are unchanged. - Sync signal sources consolidated. The live timeline only subscribes to append triggers (`onNewEvent`, `onInsert`); `onChange`, `onRemove`, and `onUpdate` had no legitimate driver in a single-user append-only sync model and are no longer wired. The `signalTimelineChange` / `signalTimelineRemove` / `signalTimelineUpdate` counters are dropped from `metricsSnapshot` and the `liveScan.summary` breakdown. - `AppLifecycleRescanObserver` removed. The 30 s wake detector in `MatrixStreamLiveScanController` and the connectivity-driven rescan already cover resume; the lifecycle observer duplicated those triggers and produced the `forceRescan.skipped (already in flight)` noise (116 of 321 lines in the 2026-04-20 log). ## [0.9.967] - 2026-04-20 ### Changed - Task detail header rebuilt against the desktop Figma as a scoped migration. Replaces the pinned title sliver + legacy `TaskHeaderMetaCard` with a new `DesktopTaskHeader`: - Three explicit lines: title → classification (category, project, labels) → metadata (due date, estimate, priority, status); each row wraps on narrow widths. - Inline capsule-style multi-line title editor (Heading 3 bold, pencil-on-hover, check / cancel on edit). ⌘S / Ctrl+S / ⌘-Enter commit; Esc cancels. - Figma priority / status / project / due-date / work-category chips; estimate chip with progress bar and overtime styling; assigned-label chips as outlined pills with a leading color dot and high-emphasis primary text. - Subdued placeholder chips for every empty state ("No project", "unassigned" category, "No due date", "Add Label") that open the corresponding picker when tapped. Long-press on a label chip still reveals its description dialog. - Due-date urgency is now tri-state: orange for due today, red for overdue, subdued otherwise. - Metadata row uses a text-scale-aware breakpoint: both groups sit side-by-side (space-between) on wide viewports and stack cleanly on narrow ones / with accessibility text scale. - The in-header `more_vert` ellipsis is gone — entry actions stay on the pinned app bar. That app bar now also surfaces the task title in `subtitle2` once the header scrolls out of view. - Presentational widget is Riverpod-free and exercised in Widgetbook (Default, Editing, Long title (wraps), Empty classification + metadata, Playground). A thin `DesktopTaskHeaderConnector` wires the existing status / priority / category / project / due-date / label pickers and `EntryController` mutations. - `TaskLabelsWrapper` is removed entirely — the estimate chip, assigned-label chips, and the Add Label affordance it used to render are all inside the header now. - AI Task Summary, Task description (agent report), Linked Tasks and Checklist cards on the task detail page now render on a flat `TaskDetailSectionCard` surface (solid `background.level02`, `radii.l`, subtle `decorative.level01` border, no gradient, no drop shadow) matching the task list. Section titles inside the touched cards use the design-system `subtitle2` token instead of Material `textTheme.titleSmall`. - Deleted: `TaskTitleHeader`, `TaskHeaderMetaCard`, and the per-chip `*_wrapper.dart` / `*_widget.dart` files that only fed the old header. Shared modal content widgets (`TaskStatusModalContent`, `showDueDatePicker`, `CategorySelectionModalContent`, `ProjectSelectionModalContent`) are retained and reused by the new connector. - Task language selection — previously a flag-shaped pill inside the header — moves into the pinned app bar's triple-dot menu as a standard list item (`ModernSetTaskLanguageItem`). The currently selected language's flag renders inline next to a "Set language" label (falling back to `Icons.language` when unset); tapping the row opens the existing `LanguageSelectionModalContent` modal and persists the selection via the journal repository. - `ActionMenuListItem` now accepts an optional `leading` widget in addition to its existing `icon`, so entries like the language action can render a flag at its natural 4:3 aspect without re-skinning the menu row. ## [0.9.966] - 2026-04-19 ### Added - Collapsible desktop navigation sidebar. A new menu icon at the top of the sidebar toggles between the expanded layout and a narrow 72 px icon-only strip. Destination labels are hidden in the narrow state and surface as tooltips on hover, while the brand logo hides to free up horizontal space. Drag-to-resize is disabled while the sidebar is collapsed so labels never end up clipped at intermediate widths; the divider still paints its hairline so adjacent panes do not shift across the transition. The collapse flag and the previously-used expanded width are persisted to `SettingsDb`, so re-expanding restores the exact prior divider position and the state survives restarts. ### Fixed - Priority / status / label pills in the task filter sheet no longer "breathe" when toggled. `_TaskFilterChoicePill` used to widen its border from 1.0 → 1.5 px on selection, which nudged every sibling chip in the `Wrap` by 1 px per edge on each toggle. The border width is now pinned at 1.5 px in every state and the pill is stateful with a 400 ms `easeInOut` animation controller that cross-fades the border and fill colour alphas between deselected and selected, so the outer dimensions of every chip are identical at every frame and neighbours never reflow. - Sync startup no longer does every step twice. The connectivity-driven `forceRescan` and the consumer's `runInitialCatchUpIfReady` used to race into `_attachCatchUp()` concurrently because neither acquired `_catchUpInFlight` for its run, producing two `catchup.waitForSync` passes, two `backfill.start events=610` fetches, and two ordered replays of the same 89-event slice (visible in logs as paired `processor.resolve` / `processor.apply` / `sequence.recordReceived` lines). `_attachCatchUp()` is now the single owner of `_catchUpInFlight` and the deferred-live-scan flush: it short-circuits with `catchup.skipped (in flight)` when another run is already in progress, and every other entry point (`forceRescan`, `runInitialCatchUpIfReady`, `runGuardedCatchUp`, `_scheduleInitialCatchUpRetry`, `startWakeCatchUp`, `_startCatchupNow`) defers to it rather than managing the flag themselves. The obsolete `bypassCatchUpInFlightCheck` parameter and `forceRescan.skippedCatchUp` log are removed. - Pre-history gap handling no longer logs per incoming event. Hosts that carry a permanent unresolved prefix (e.g. `lastSeen=81982` stuck while the VC counter keeps climbing) used to re-emit `largeGapDetected gapSize=7344`, `gapDetectedRange`, `recordReceivedEntry detected N gaps`, and `apply.agentEntity.gapsDetected count=7344` for every single catch-up event, dominating the desktop sync log. The sequence service now suppresses those lines on incremental extensions of an already materialised range and returns only the newly materialised subrange in its `gaps` result, so the caller-side `apply.*.gapsDetected` log is driven by actual new signal. A fresh idle restart now logs on the order of tens of lines instead of 400 KB. - Redundant `start.catchUpRetry` catch-up on every startup. The signal binder's 150 ms follow-up `runGuardedCatchUp('start.catchUpRetry')` now only runs when the initial catch-up has not yet completed. Once the consumer reports `catchup.initial.completed`, the retry is skipped instead of kicking a second full `_waitForSyncCompletion` + event-replay pass. - Connectivity bootstrap no longer duplicates the startup rescan. `connectivity_plus` emits the current state synchronously on subscribe, which used to schedule a `forceRescan` that raced the dedicated startup rescan. The listener now swallows the first emission (`service.forceRescan.connectivity.bootstrapSkipped`) while still recording the connectivity signal for metrics. ## [0.9.965] - 2026-04-19 ### Fixed - DailyOS mobile page no longer surfaces "Failed to load timeline" and "Failed to load budgets" on devices where the v39 migration's partial expression index `idx_journal_tasks_due_open` is missing or cannot satisfy SQLite's `INDEXED BY` clause. `JournalDb._selectTasksDue` now catches `SQLITE_ERROR` from the pinned fast path and retries once without the pin, so the open-task due-date query still returns rows when the planner cannot prove the partial's WHERE. In addition, `beforeOpen` recreates `idx_journal_tasks_due_open` and `idx_journal_task_status_private` as `CREATE INDEX IF NOT EXISTS` on every launch (guarded on the `journal` table existing), so devices that landed at `user_version = 39` with a missing index self-heal on next open. The CREATE statements now live in a single pair of private top-level constants reused by `onUpgrade`, `beforeOpen`, and the migration tests. Regression tests cover both the self-heal path and the unpinned fallback in `test/database/task_indexes_v39_migration_test.dart`. ## [0.9.964] - 2026-04-19 ### Changed - Outbox / sync DB: removed a 40–600 ms main-isolate blocking query that was triggered on every outbox enqueue (including each image paste or drag-and-drop). `getLastSentCounterForEntry` used to compute `MAX(counter)` against an index that did not include `counter`, forcing SQLite to read every matching row from the heap for hot entry_ids. The v11 schema replaces `idx_sync_sequence_log_host_entry_status` with a covering `idx_sync_sequence_log_host_entry_status_counter`, the query is rewritten as `ORDER BY counter DESC LIMIT 1` (index-only scan with early terminate), and the sequence service now caches the last-sent counter per `(hostId, entryId)` with LRU capacity 2048. Follow-up lookups for the same entry hit the cache and skip the DB entirely; the cache is kept consistent by refreshing it on `recordSentEntry`. Under the same load as 2026-04-19 (3092 invocations, p50 43 ms, peak 618 ms) this collapses to a single sub-millisecond index probe per distinct entry per TTL window. Removes the multi-second UI freeze observed when pasting/dragging images. - Sync: eliminated the dominant source of redundant work on devices that carry a permanent pre-history gap in the sequence log. Previously, every incoming event on such a host re-ran a multi-chunk scan of the `sync_sequence_log` table plus a batch-insert pass over thousands of already-materialized rows, and emitted four INFO log lines per event. The sequence log now memoizes the highest counter already materialized as a missing range per host and short-circuits the entire pass when the observed range is already covered. When the observed counter does advance, only the incremental delta is scanned and logged. This removes the main cause of severe mobile UI blocking on slow sync queries and the bulk of the daily sync-log volume on desktop. - Sync: missing or requested rows whose backfill request count reaches the per-entry cap are now retired to the terminal `unresolvable` status automatically at the start of each backfill cycle. Without this, a permanently unresolvable counter — a pre-history entry, a purged payload, or a mapping whose payload vector clock is permanently behind the requested counter — kept the contiguous-prefix watermark in `getLastCounterForHost` stuck forever, which in turn forced gap detection to re-enter on every subsequent event for the same host. The watermark now advances past these rows and gap detection stays quiet. - Sync attachments: the matrix SDK's "File is no longer cached" error used to be emitted as a full exception with stack trace on every replay of an evicted attachment event, with zero recovery value. The ingestor now records the event id the first time this happens and short-circuits subsequent attempts immediately, collapsing bursts of hundreds of error entries per day to one info line per eviction. - Sync: duplicate in-flight descriptor fetches for the same attachment event are now deduplicated. Two text events referencing the same `jsonPath` within a single catch-up or live-scan wave previously each launched an independent download, decrypt, and decode; they now share a single future keyed by the index key and descriptor event id. - Outbox: the `sendNext.state` and `dbNudge count=...` observability lines are now coalesced. They emit on state transitions and at a 30-second minimum interval instead of once per tick, removing the majority of steady-state outbox log lines without changing enqueue or send behavior. ## [0.9.963] - 2026-04-19 ### Added - Agent suggestion panel now surfaces the reasoning behind resolved proposals without cluttering the list. A collapsed "Recent activity" strip shows the three most recent ledger entries (newest-first) with verdict icons — check for confirmed, X for rejected, undo for agent-retracted — and when a rejection or retraction carries a reason it lives behind an info icon whose tooltip reveals the text on tap or hover. The strip stays hidden entirely when there is no resolved activity to show. A "Confirm all" action sits next to the pending badge whenever more than one open suggestion is pending, fanning out to `ChangeSetConfirmationService.confirmAll` per distinct change set. - Initial-language auto-apply: `set_task_language` now bypasses the deferred-confirmation flow when the task has no language set yet, so the agent can pick the language silently on first use. Re-tagging a task that already has a language continues to flow through the standard change-set approval path, and the same single-use guard that protects the title carve-out prevents a second call in the same wake from silently overwriting the freshly-applied language. ### Changed - Suggestion rows no longer display the raw snake_case tool key beneath the human summary. The tool name was developer-facing noise that did not help the user decide whether to confirm or reject a proposal. ## [0.9.962] - 2026-04-19 ### Changed - Task agent: the mandatory `update_report` tool call at the end of every wake is now a hard guarantee instead of a prompt-only request. When the model stops without calling it — a routine failure mode for weaker local models such as Qwen 3.6 served via `mlx-vlm` — the workflow now issues one additional inference pass with `tool_choice` pinned to `update_report` and a direct reminder message, forcing the final report. `ConversationRepository.sendMessage` and the underlying inference layer (`CloudInferenceRepository`, `CloudInferenceWrapper`, `InferenceRepositoryInterface`, `OllamaInferenceRepository`) gained a threaded `toolChoice` parameter that overrides the default `auto` selection policy; the override is currently honored on the OpenAI- compatible path and silently ignored by the Gemini/Ollama/Mistral sub-repositories, which still benefit from the retry's pointed user message. The task-agent scaffold and the seeded report directive also now lead with a non-negotiable "final step" instruction so compliant models need no retry at all. ### Added - Initial-title auto-apply: when the task agent calls `set_task_title` on a task whose title is still null or empty, the title is applied immediately without a user-approval prompt, so a freshly dictated task lands with a meaningful name instead of an empty-looking suggestion waiting in the panel. Once a title is present the tool reverts to the standard deferred confirmation flow, and a second silent-abort guard inside `TaskToolDispatcher` no-ops any race in which a manual edit, synced edit, or previously-confirmed proposal populated the title between the LLM emitting the call and the dispatcher running. - Task agent suggestions are now surfaced in a single consolidated panel on the task detail page instead of two disconnected cards, and the agent itself can autonomously retract its own stale proposals. On every wake the agent receives a unified proposal ledger listing every suggestion it has ever made for the task, grouped into open vs. resolved entries and carrying stable fingerprints. A new immediate `retract_suggestions` tool lets the agent withdraw proposals that are no longer relevant (the task state already matches them, the user made the change manually, or they duplicate another open proposal) without any user prompt — the item simply disappears from the active suggestion list, leaving an auditable trail alongside user confirmations and rejections. Dedup now excludes retracted items from its basis so the agent can cleanly re-propose after the task context materially changes, while user rejections remain sticky. ## [0.9.961] - 2026-04-18 ### Changed - Sync pipeline: removed duplicate-work in inbox attachment handling and sync-family logging. `AttachmentIndex.record` now dedupes per eventId so repeated observations from live scan + catch-up + backfill passes become no-ops instead of thrashing the per-path slot between events sharing one `relativePath`. `SyncEventProcessor` and `SyncSequenceLogService` no longer emit each log line twice (once via `DomainLogger`, once via a paired direct `LoggingService.captureEvent` to a sync-file domain) — `DomainLogger` is the single emitter and falls back to a direct `sync`-domain capture when no domain logger is wired. Descriptor catch-up skips events whose `relativePath` is not in the pending set, so the per-run scan no longer records ~1000 unrelated events into the attachment index. Removed the per-tick `OUTBOX enqueueRequest() done` log, which fired on every debounced wake and carried no signal. - Sync pipeline: the per-event apply loop in `MatrixStreamProcessor` now runs inside a single `JournalDb.transaction`, so a slice of N sync events commits once instead of N times. Drift coalesces the journal table stream emissions to one notification per slice, collapsing the reader/writer lock contention that produced multi-second waits behind otherwise-cheap `id IN (?)` lookups during large catch-up replays. Per-event error handling is unchanged: individual apply failures are still caught locally and scheduled for retry, so the transaction only rolls back when a commit genuinely fails. ## [0.9.959] - 2026-04-18 ### Changed - Desktop task switcher: switching between open tasks in the desktop split layout now crossfades the detail pane instead of cutting abruptly. The transition is a 480 ms `easeInOutCubic` fade driven by an `AnimatedSwitcher` keyed on the selected task id, so it fires only when a genuinely different task is opened — reloads and in-place data updates for the currently open task do not trigger the animation. ## [0.9.958] - 2026-04-18 ### Changed - Desktop navigation sidebar: the Tasks and Settings notification counts now render as a design-system number badge in a dedicated trailing slot on the right of each row (instead of a Material `Badge` overlapping the icon and label). The vertical spacing between sidebar items is now 24 px to match the Figma layout. Sidebar rows are no longer pinned to a fixed 48 px height — each row grows with its label, so increasing the OS/app text scale no longer vertically clips "Projects", "DailyOS", "Insights", "Logbook" or "Settings". The mobile bottom-nav preserves its existing overlay badge. ## [0.9.957] - 2026-04-18 ### Changed - Database schema bumped to v39 with two new partial indexes on the `journal` table: `idx_journal_tasks_due_open` (keyed on the JSON-extracted due date for open tasks) lets the due-date query stream straight from the index without an external sort, and `idx_journal_task_status_private` covers `countInProgressTasks` and similar global task-status counts with a narrow index instead of scanning the full task set. Both partials carry `type = 'Task' AND task = 1 AND deleted = FALSE` so they stay consistent with the other task indexes and SQLite's theorem prover reliably proves coverage for the paired queries. ## [0.9.956] - 2026-04-18 ### Fixed - AI-proposed time entries can now be confirmed from the task "Proposed changes" panel. The approval-time validator previously rejected any `create_time_entry` whose `startTime` or `endTime` fell after the originating wake timestamp, which left suggestions like 11:13–11:48 or 11:00–13:00 stuck as pending with a "Failed to apply change" toast whenever the agent rounded or estimated a session past the wake instant. The wake timestamp is still used for the same-day check (so after-midnight approvals keep working), but the "not in the future" cutoff no longer applies at approval time — the user is the authority when confirming. `endTime` must still land on the same day as `startTime`. Direct agent-time tool calls (if ever wired) still refuse fabricated future times. ## [0.9.955] - 2026-04-17 ### Changed - Align the desktop tasks and projects surfaces with the Figma design system. Sidebar, task list, task details and projects list now share the deeper `background/01` surface, and task cards sit on `background/02` so the subtle card/raised-surface contrast matches the Figma reference. Dark-mode `text/*` tokens are raised to the Figma-specified opacities (high 100%, medium 80%, low 64%) so headlines, subtitles and timestamps read with the correct contrast without changing type weights. - The selected task row uses the Figma `surface/selected` token (teal at 16%) for a subtler highlight instead of the previous stronger tint. - Task group headers ("Today", "Yesterday" …) and the `N tasks` counter use the caption (12 px) token for a lighter hierarchy. - Task title header and the pinned `SliverAppBar`s above it render on `background/01` with no elevation, so the strip no longer reads darker than the body. - Sidebar drops the legacy "+ New" quick action, always uses the filled settings glyph, and swaps the Ionicons/MDI nav glyphs for Material rounded/outlined equivalents (Tasks, Projects, DailyOS, Habits, Insights, Journal, Settings). The Insights nav tab now reads "Insights" (English/en-GB) and "Einblicke" (German). - Task list filter chips use the design-system chip component and render inline under the search bar, with priority chips carrying a leading priority glyph. Tasks and projects filter triggers now use `filter_list_rounded` to match the Figma funnel glyph. - Tasks and projects headers are pinned outside the scroll view so the title stays static and pull-to-refresh only drags the list below it. - The floating action button component is now a rounded-24 teal surface (`radii.xl`) matching Figma, and the journal/task details FAB was swapped from Flutter's default circular FAB to the design-system FAB so list and details FABs share the same shape and bottom-right anchor. The task details showcase action bar uses the Figma `subdirectory_arrow_right` glyph for the "linked" round action. ## [0.9.954] - 2026-04-17 ### Added - Optional gzip compression for JSON sync attachments, gated by the `use_compressed_json_attachments` config flag (off by default). Receivers unconditionally decompress attachment events that carry a `com.lotti.encoding: gzip` marker, so peers already on this release stay compatible once the flag is flipped on. Only `.json` attachment paths are compressed; media files are already on compressed formats and skip the branch. - The new compression flag is toggleable from Settings → Flags, with localized labels and descriptions across all supported languages. ## [0.9.953] - 2026-04-16 ### Fixed - Raise the open file descriptor soft limit to 10,240 at startup on macOS and Linux. Apps launched from Finder/Spotlight on macOS inherited launchd's legacy 256 limit, which was trivially exhausted under real-world load (sockets, SQLite handles, attachment writes, log files). Under slow-network conditions this manifested as sporadic `EMFILE` errors in the Matrix sync pipeline and cascading DNS lookup failures until the app was restarted. ### Added - Log the file descriptor soft/hard limits and the adjustment outcome at startup (`MAIN / fdLimits`), so the current ceiling is discoverable from the app logs without requiring `launchctl` or `ulimit` inspection. - Annotate `EMFILE` failures on the attachment save path with current descriptor limits (`attachment.save.emfile` event), making future FD pressure incidents self-diagnosing. ## [0.9.952] - 2026-04-15 ### Changed - Task filter modal redesigned with the design system filter sheet, replacing the old Material chips and switches with pill-shaped sort/priority selectors, tappable selection fields for status/category/label, and Clear All / Apply action buttons. - All filter modals (task, project, selection sub-modals) now use Wolt modal sheet for consistent adaptive presentation (bottom sheet on mobile, dialog on desktop). - Selected filter chips now show contextual icons: status icons with color, category icons from the category definition, and colored dots for labels. - Priority glyphs updated to match Figma: P0 uses new_releases icon, P1 uses signal_cellular_alt, P2/P3 use ascending bar variants with faded unfilled bars. Priority pill spacing tightened so all options fit on one row. - Categories and labels list pages now use design system list items in a grouped container, matching the visual style of the settings page. - Categories page header replaced with the standard SettingsPageHeader for consistent back-button behavior across mobile and desktop layouts. - Dashboards, measurables, and habits list pages now use design system list items in a grouped container, consistent with categories and labels. - Maintenance and logging domain settings pages now use design system list items in a grouped container, consistent with the rest of the settings pages. - Sync settings and sync maintenance pages now use design system list items in a grouped container, consistent with the rest of the settings pages. - Settings page uses a 3-column desktop layout for wider screens. ## [0.9.951] - 2026-04-13 ### Added - Project one-liner: AI-generated summary subtitle on project list cards, giving an at-a-glance status for each project (mirrors the existing task one-liner feature). - Project agent now produces a dedicated `one_liner` field in its report, separate from the longer TLDR summary. ## [0.9.950] - 2026-04-11 ### Removed - Removed the gamey theme and all gamey-styled widgets (GameyFab, GameyCard, GameySettingsCard, GameyJournalCard, GameyTaskCard, shimmer/celebration overlays). Standard Flutter widgets and existing modern cards are used instead. ### Added - Resizable panes in desktop layout: users can drag dividers between the navigation sidebar and content area, and between the list and detail panes, to adjust widths. Pane sizes persist across sessions via SettingsDb. ### Fixed - Tasks desktop split view now keeps the current loaded list window mounted until refreshed pages resolve, which reduces visible flicker during task saves and other live updates while preserving regrouping after sort changes. ## [0.9.949] - 2026-04-10 ### Changed - Redesigned dashboard list page: removed the old SliverAppBar top bar, replaced flat list items with rounded grouped cards using DesignSystemListItem, and applied design system tokens for spacing, colors, and typography. ### Added - AI one-liner subtitle in task list: each task row now displays the agent-generated summary below the title (max 2 lines), giving users at-a-glance progress visibility without opening the detail view. - Time recording indicator in task list: a red dot appears on any task row that is actively recording time, positioned as a trailing element in the row layout. - Live task data in list rows: editing a task's title, status, or other data in the detail pane now updates the corresponding list row in-place via a lightweight `taskLiveDataProvider`, without requiring a full list refresh. ### Fixed - Removed animated transitions (`animateTransitions`) from the paged task list to eliminate visual flicker when loading pages or refreshing data. ## [0.9.948] - 2026-04-10 ### Changed - Task filter modal: replaced the old custom filter modal with the design system version from Widgetbook. The new modal uses the same visual language as the projects filter and includes sort, status, priority, category, label, project, and agent assignment filters with display toggles. - Project filtering now lives inside the task filter modal with a grouped selection view (projects grouped by category). Removed the separate ProjectHealthHeader from the task list. - Search mode (full text vs. vector) is now only shown on desktop when vector search is enabled. On mobile, full-text search is always used. - Cover art thumbnails are now always shown on task cards. ### Removed - "Show projects header" display toggle (superseded by in-modal project filter). - "Show distance on cards" display toggle (was a debugging tool only). - "Show cover art" display toggle (cover art is now always shown). ### Added - Checklist filter tabs: added a "Done" tab alongside "Open" and "All" to filter checklist items by completion state. Tab order is Open, Done, All. - Desktop mode: responsive layout switches from mobile bottom navigation to a persistent left sidebar when the window is wider than 960px. The sidebar includes all navigation destinations with Settings pinned at the bottom, plus the "New" and AI assistant action buttons. - Desktop split-pane view for Tasks, Projects, and Dashboards: the list and detail views are shown side by side. Selecting an item in the list reactively opens its detail in the right pane without losing list scroll position. Full-width sections (Habits, My Daily, Journal, Settings) continue to use the entire content area. - Active item highlighting: the selected task or project card is visually marked in the list pane when a detail view is open. - Localized empty-state placeholders for the detail pane in all seven languages. ### Fixed - Checklist cards no longer play a visible collapse/expand animation on first render. Items snap into their initial state immediately, preventing jank when multiple checklists load at once on slower devices. ## [0.9.947] - 2026-04-09 ### Changed - Bottom navigation rollout: every main app tab now uses the Widgetbook design-system bottom navigation shell instead of the legacy Spotify-style bar, with the floating pill centered consistently across the app. - Mobile bottom sheets and Wolt modals now open on the root navigator on narrow screens so they render above the floating bottom navigation and block tab switching while a modal is active. - Tasks and Projects now share the same design-system floating action button, and the shared bottom-navigation spacing now lifts floating action buttons and recording indicators clear of the navigation chrome across list, detail, and settings surfaces. - Task status selection modal redesigned: replaced wrap-of-chips with a vertical list of tappable rows — each showing the status icon, label, and a trailing checkmark on the currently-selected status. - Status icons updated to match design reference: Groomed uses a pencil icon, In Progress a play arrow, Blocked a warning triangle, On Hold pause bars. - Status icon colors are now brightness-aware: lighter muted shades on dark backgrounds, higher-contrast shades on light backgrounds. - Icon mapping centralised in `taskIconFromStatusString` so the status chip and the selection modal always stay in sync. - Status rows wrapped with `Semantics` for screen-reader accessibility. ## [0.9.946] - 2026-04-08 ### Changed - Checklist widget redesigned to match Widgetbook reference: progress ring now always visible in the header (collapsed and expanded), filter strip (Open/All) uses a highlighted background for the selected tab, and the add-item field is a plain text field without suffix icons. - Checklist item rows no longer enter edit mode on title tap; only the pencil icon opens the inline editor. - AI suggestion dialog strings in the checklist item row are now fully localized. - "Share" label in the checklist card menu is now localized. - Fixed `ChecklistItem` rendering in the entry details view. - Progress ring and completion fraction are now hidden when a checklist has no items, and the three-dot menu is hidden when no actions are available. - Checklist item row now uses a minimum height instead of a fixed height so the inline editor can expand for multi-line text and content is no longer clipped when the system text scale is increased. - Add-item field uses a minimum height constraint instead of a fixed height. - Increased spacing between the progress ring and the completion fraction text. ## [0.9.945] - 2026-04-08 ### Added - Standalone soul evolution: dedicated 1-on-1 sessions focused on personality refinement, aggregating feedback from all templates sharing a soul. Accessible via "Soul 1-on-1" on the soul detail page, with its own review page and session history. ### Changed - Tasks tab migration: added a new design-system tasks browse page behind the `enable_tasks_redesign` runtime flag, wired through a `/tasks` root switch so the legacy page stays available during rollout. The new page keeps the existing infinite-scroll paging controller, current filter model and modal, create-task flow, and task-detail navigation while rendering sort-aware section groupings for due-date, priority, and creation-date presentations. - Tasks tab redesign: aligned the grouped-task card presentation more closely with Figma by switching to the actual status glyphs, inset dividers, grouped header treatment for priority sections, tracked-time metadata, and a filter modal search-mode control instead of an inline full-text/vector toggle. - Tasks tab polish: replaced the provisional priority glyphs with the actual Figma P0-P3 icons and restored the intended breathing room below the search row. - Tasks and projects grouped cards: both tabs now share the same hover and selection row-surface behavior so active rows blend over adjacent dividers consistently, while task category chips now use the same category tag component as projects and the task rows have slightly looser spacing. - Bottom navigation migration: the redesigned Tasks tab and the Projects tab now render the Widgetbook design-system bottom navigation bar, while the remaining tabs keep the legacy Spotify-style navigation during the rollout. ## [0.9.944] - 2026-04-08 ### Added - Desktop zoom: Cmd+/- keyboard shortcuts and View menu items for zooming the app UI in and out, with Cmd+0 to reset. Zoom level persists across sessions. ## [0.9.943] - 2026-04-07 ### Added - Token usage Stats tab in Agents Settings — iOS battery-usage-inspired dashboard showing 7-day and 30-day token consumption with interactive bar charts, average vs. today comparison, per-model breakdowns, and per-template source activity with high-usage alerts. - Wake activity chart migrated into the new Stats tab for a unified view. ## [0.9.942] - 2026-04-07 ### Added - Soul evolution UI: standalone soul evolution page with an interactive evolution catalog, soul selector, and a conversation-based refinement workflow that lets users iteratively shape agent soul documents from templates. - Pluggable soul document: soul documents can now be customized per agent template, decoupling soul content from the shared default. - One-on-one v2: redesigned agent one-on-one conversation experience. - Manual planning flow: new "Set time blocks" step for manual daily planning. - Customer care dashboard: new frontend with admin authentication. ### Changed - Proposed changes card: removed inline confirm/reject buttons from individual change items; users now confirm or reject via swipe gestures or the bulk "Confirm all" button. - Settings page restyled to use design system list items, matching the current Figma visual language. - Context menu component aligned with Figma design specifications. - Projects tab: task rows now display agent one-liner summaries, follow-up tasks automatically inherit the parent project, and the layout alignment and scroll behaviour have been improved. - Pending agent wakes UX: improved wake card presentation and prevented redundant wake scheduling. ### Fixed - Agents no longer trigger self-notification loops. - Projects page no longer flickers during live data refreshes. ## [0.9.941] - 2026-04-02 ### Changed - Agents settings: added a `Pending Wakes` dashboard under `Settings > Agents` with a live badge count, per-wake countdown cards, and per-card wake deletion for deferred and scheduled wakes. - Agents settings: the `Instances` tab now shows a compact task-agent fleet summary under the filters, with current totals for all, active, dormant, and destroyed task agents. - Task-agent wake prompts no longer inject sibling-task TLDR directories, and the related-task drill-down tool now stays defined in the registry while being hidden from the LLM until a stronger retrieval path is in place. ## [0.9.938] - 2026-03-29 ### Changed - Projects detail task summaries: project-task rows now show the latest task-agent `oneLiner` subtitle between the task title and metadata, matching the current Figma spacing and typography on mobile and desktop layouts, and open project detail pages now refresh automatically when a task-agent report finishes. - Task-agent reports: the `update_report` contract now requires a dedicated `oneLiner` tagline alongside the TLDR and full report so task cards can show a concise current-state summary without reusing the longer report text. - Projects detail data path: highlighted task rows now bulk-load the latest task-agent reports for all linked task IDs in one repository call, avoiding N+1 report queries on larger projects. - Projects list grouped cards: category group containers now render the same subtle bordered card treatment shown in the current Figma list designs, instead of relying on background fill alone. - Projects detail live refresh: when related-task updates reload the derived project detail record, the page now keeps the previous content mounted until the new record resolves, avoiding full-page flicker and scroll-position jumps. - Projects detail scrolling: the top-level detail route now renders the task panel through a sliver-based lazy list, which keeps large project task lists steadier during fast downward scrolling instead of eagerly laying out every task row. ## [0.9.937] - 2026-03-29 ### Changed - Follow-up tasks now automatically inherit the project of the parent task they were created from, both from the UI (create task button, linked tasks header, desktop menu) and from the agent follow-up task workflow. ## [0.9.936] - 2026-03-28 ### Changed - Design system navigation: the mobile detail back control is now a reusable component with the intended regular-weight label styling, so detail pages do not drift into heavier back-button text. - Projects detail task statuses: project-task rows now use the Figma-aligned colored status glyphs for in-progress, done, on-hold, blocked, and related task states instead of flattening every status icon to the same gray tone. - Projects detail task typography: project-task titles are explicitly kept at regular body-small weight, and the duration plus status line now uses one shared compact caption text style. - Projects detail task list: task rows now let long task titles wrap across multiple lines and keep duration plus state metadata underneath, with extra vertical breathing room in the `Project Tasks` header. - Projects list spacing: widened the grouped project-row horizontal inset so titles, metadata, and status pills sit with the intended breathing room inside each card. - Projects detail redesign: the top-level Projects flow now opens a new design-system-aligned detail page at `/projects/:projectId`, with editable category, due date, and status controls, a health-band header, manual agent refresh, an expandable AI report, and the current project-tasks card. - Projects detail data path: the new detail surface now builds from live project, task, category, and project-agent state while reusing the same shared component tree in the app and Widgetbook. Project-task rows now open task details directly, while the older settings-scoped project views remain in place for category-driven flows during the rollout. - Projects detail UX polish: category and due-date pickers are now reachable even when no value is set, the empty-state report copy only appears when both summary and full content are absent, and expanded reports strip the leading H1 title that the UI already renders above the report section. - Projects visual polish: grouped project rows, progress rings, and project task rows now match the linked Figma spacing more closely, with progress colors driven by completion level and the mobile detail title reduced to the intended heading size. - Projects tab search: the live top-level Projects page now enables local substring search against the grouped list, and the shared design-system search field keeps typed text stable across rebuilds while centering the search text correctly on mobile and desktop surfaces. - Task agents now receive parent-project context on wake when a task belongs to a project, including the latest project-agent TLDR and full report body so task recommendations can align with project-level priorities and blockers. - Task agents now also receive a bounded directory of related tasks from the same project, using stored sibling task-agent TLDRs plus logged time spent, and can inspect one of those sibling tasks on demand through a new read-only drill-down tool scoped to the current wake context. ## [0.9.933] - 2026-03-27 ### Changed - Projects tab: the new top-level Projects tab now ships behind the `enableProjects` feature flag with the intended design-system mobile layout. The live tab and the Widgetbook mobile reference now mount the same sliver-driven header, grouped category sections, and floating create action, while the search field stays disabled pending the later vector-search follow-up. - Projects filters: the Projects tab and Widgetbook showcase now share the same design-system filter modal, nested field-selection sheet, and state model for project-status and category filtering instead of keeping separate mock-only filter behavior. - Projects tab reactivity: grouped project snapshots now refresh automatically when relevant project, task, category, or private-visibility updates arrive, so status changes such as `Active` to `Completed` are reflected in the UI without reopening the page. ## [0.9.932] - 2026-03-23 ### Changed - Projects now expose a user-facing health band (`Surviving`, `On Track`, `Watch`, `At Risk`, `Blocked`) in the tasks-page project header and project detail page based on the latest project-agent report. - Project health tracking now marks project summaries stale when any linked task activity changes, and the app state layer now aggregates health bands, stale-summary state, and active project recommendations for future dashboard surfaces. - Direct edits to a project and task linking changes now use the short deferred wake path, while other task-level activity stays stale-only until the next scheduled project digest unless refreshed manually sooner. - Projects now also have a top-level tab behind the `enableProjects` feature flag, reusing the existing bottom navigation and loading grouped project/category data through a single batch overview path. ## [0.9.931] - 2026-03-23 ### Added - Qwen 3.5 Plus model: added Alibaba Cloud's advanced multimodal reasoning model with text and image understanding for complex analytical tasks. - Chinese AI Profile: new default inference profile using Qwen 3.5 Plus for reasoning alongside Qwen VL Flash (vision), Qwen Omni Flash (audio), and Wan 2.6 (image generation). - Alibaba FTUE setup dialog: added first-time setup wizard for Alibaba Cloud providers with model and category creation. ### Changed - FTUE setup simplified: provider setup wizards no longer create prompts — all AI capabilities are handled by the skill-based automation system via inference profiles. - Project agents now skip dormant daily digests, track pending project activity in synced agent state, and show stale summaries in the tasks-page project header until the next 06:00 report refresh. - Project detail pages now render the current project-agent report inline, expose a manual refresh action, show when no project agent exists yet, and return to the originating category instead of dropping back to the top settings page. - Confirmed project-agent next-step proposals now persist as active project recommendations that supersede older active guidance and can be resolved or dismissed from the project detail page. ## [0.9.929] - 2026-03-21 ### Added - Design system Tabs: added a new Widgetbook-first Tabs component with small/default sizes, token-driven interactive states, optional counter and icon slots, and focused widget/widgetbook coverage. - Design system Calendar picker: added a Widgetbook-first calendar picker with month rail navigation, token-driven monthly range cells and weekly date cards, plus focused widget/widgetbook coverage. - Design system Progress bar: added a Widgetbook-first determinate progress bar with default and chunky variants, optional header content, token-driven fill and track treatments, and focused widget/widgetbook coverage. ### Changed - Design system controls: tightened accessibility contracts for unlabeled controls, added real tooltip behavior and resilient label truncation to toggles, reset transient interaction states when controls become disabled, added a disabled state to split buttons, and now enforce the same visible-label-or-semantics contract across chips, tabs, dropdowns, and split buttons. ## [0.9.928] - 2026-03-17 ### Added - Voice-driven time tracking: the task agent can now create time tracking entries from voice dictation. Users can say things like "I worked on the API from 2 to 4 PM" or "Start a timer at 7 PM" and the agent will propose a journal entry with the correct timestamps. Supports both completed sessions (start + end) and running timers (start only). Entries require user approval via the change set confirmation UI. - Projects UI: full project management interface with create/detail pages, project picker for tasks, category integration, health header on tasks page, and project-based task filtering. Each project auto-creates an AI agent for analysis and reporting. - Projects feature flag: all project UI is gated behind the `enableProjects` config flag (default: off), keeping the feature hidden until ready for release. - Expandable project health header: the tasks page header now shows a collapsible summary card with project/task counts that expands to per-project rows with status chips and navigation. - Project status picker: interactive bottom sheet on the project detail page for changing project status across all five variants. ### Removed - Prompt tab and prompt editing page from AI settings. Prompts are now managed internally and no longer require manual user configuration. ### Removed - Tags: removed the entire tag concept — generic tags, person tags, and story tags — including all UI (tag modals, tag settings pages, tag chips on entries), sync/outbox infrastructure, services, and repositories. The `tag_entities` and `tagged` database tables are intentionally left in place to avoid migrations but are no longer accessed or defined in the schema. Deprecated fields are retained in data models for JSON backward compatibility. ## [0.9.927] - 2026-03-16 ### Fixed - Matrix sync catch-up: fixed two bugs that prevented cold-restart catch-up from paginating server history correctly. Backfill no longer short-circuits on marker ID before reaching the timestamp boundary, and the local expansion loop no longer overwrites paginated events with a smaller local-only snapshot. ## [0.9.926] - 2026-03-15 ### Added - Category AI defaults: categories can now define a default inference profile and agent template. New tasks created in the category automatically inherit the profile (for speech-to-text and image analysis) and get an agent auto-assigned from the template. - Agent content-gating: auto-assigned agents enter a "waiting" state and only activate once the task has meaningful content (linked entries with non-empty text), preventing premature agent runs on blank tasks. - Task-level profile fallback: when a task has no agent, the profile automation system falls back to the task's inherited profile for skill automation (transcription, image analysis). ## [0.9.925] - 2026-03-15 ### Fixed - Notifications: honor the notification config flag across all notification methods. Previously, the "tasks in progress" badge notification and general show notifications were sent even when notifications were disabled. ## [0.9.924] - 2026-03-15 ### Added - Cover art skill: image generation now runs as a background skill via the unified skill pipeline, with fire-and-forget workflow that auto-imports the generated image and assigns it as task cover art. - Reference image selection: linked-task cover art images are now discovered automatically via bidirectional task graph traversal (with link icon overlay), and the selection limit has been increased from 3 to 5 images. - Cover art modal: stays open after triggering generation with Siri waveform progress animation; dismissible without stopping background generation. - Automatic image analysis: newly generated cover art images automatically trigger the image analysis skill, just like a manual photo drop. ## [0.9.923] - 2026-03-15 ### Added - Inference profiles: new "Thinking (High-End)" model slot for tasks where quality matters more than speed/cost (e.g. coding prompt generation). Falls back to the regular thinking model when not configured. ## [0.9.922] - 2026-03-15 ### Changed - AI popup menu: skills now appear in a dedicated top section above legacy prompts, with two-section layout and profile-based skill triggering. - Automatic AI processing: only profile-driven automation runs automatically for audio transcription and image analysis; the legacy category-based automatic prompt fallback has been removed. - Category details UI: removed the "Automatic Prompts" configuration section (data remains in DB for backward compatibility). - Skill inference status: SkillInferenceRunner now updates InferenceStatusController so the Siri waveform animation shows during skill-driven inference. - AI popup visibility: hasAvailablePromptsProvider now considers both skills and legacy prompts when deciding whether to show the AI assistant button. - Preconfigured prompts: stopped seeding `audio_transcription` and `audio_transcription_task_context` legacy ASR prompts. ## [0.9.921] - 2026-03-14 ### Changed - Sync catch-up: reconnect recovery is now timestamp-first, paging backward until the stored timestamp boundary is visible, then replaying forward with bounded overlap. Local echo ids are no longer persisted as durable remote markers, and history requests now honor the configured page size. - Sync reliability: backfill nudges wait for ordered replay batches to finish, and incomplete catch-up keeps live scans deferred instead of treating partial recovery as ready. - Database diagnostics: all Drift-backed databases now log slow queries into daily `slow_queries-YYYY-MM-DD.log` files when the logging domain is enabled. - Database indexes: new dedicated indexes for task due-dates, date-sorted task lists, journal browse, definition lists, linked-entry lookups, sync sequence scans, outbox queue scans, and agent database thread/saga/template queries. Index-adding migrations now safely recreate existing indexes to prevent startup failures. - Journal query improvements: common browse queries use simplified fast-path SQL, task filters use `EXISTS` predicates instead of `CASE` branches, linked-child lookups use direct joins, and all-starred/all-private paths bypass redundant filters. Empty selections short-circuit before hitting SQLite, and bulk id-based fetches use an unsorted path when order is not needed. - Task progress: task-card progress lookups now batch estimates and linked work time spans across visible task IDs through lightweight queries instead of loading full entities. - Settings database: reads and writes now use a direct Drift executor instead of a background-isolate hop, and same-value saves return early. Settings-key lookups are cached in-process with concurrent cold reads coalesced. - Config flags: `JournalDb` keeps a shared in-memory config-flag snapshot so per-flag consumers no longer re-query SQLite after bootstrap. Startup shell and navigation listeners watch only the flags they need. - AI configs: type-based config watchers derive from a shared in-memory snapshot instead of re-querying `ai_config.sqlite`, and label-definition visibility checks use explicit private-status filters. - Daily OS: the unified day view now ignores unrelated database notifications instead of reloading its full bundle on every update. ## [0.9.919] - 2026-03-12 ### Changed - Sync inbox: repeated replay of the exact same attachment event no longer re-runs attachment observe/download/write work unless the local file is missing or empty. This reduces sync log volume and unnecessary agent payload I/O during large catch-up waves. - Sync logging: sync-family info logs are now routed into `sync-YYYY-MM-DD.log` instead of inflating the general `lotti-YYYY-MM-DD.log`. Sync errors remain visible in both logs. - Sync diagnostics: receiver-side signal logging now emits burst/pass summaries with source breakdowns instead of one hot-path line per raw scheduler poke. - Sync backfill follow-up: re-request paging now walks past already queued oldest rows, in-flight `sending` backfill requests suppress duplicates, and zombie-file cleanup rejects paths that resolve outside the docs directory. - Sync reliability: marker-missing catch-up no longer replays a bounded recent tail as successful backlog, host progress now stays on the highest contiguous resolved counter, large gaps are fully materialized for backfill, and newly detected missing work nudges automatic backfill immediately. - Sync convergence: reconnect catch-up now keeps paging until it either finds the stored Matrix marker or reaches the stored timestamp boundary, then replays that ordered historical slice instead of turning an ordinary offline backlog into redundant backfill requests. - Agent sync startup: the app now keeps agent initialization alive from app startup when agents are enabled, so incoming agent backfill can resolve without waiting for the first entry or agent screen to be opened. - Session rating: replaced the intrusive automatic rating modal (which popped up when a timer stopped) with a non-blocking pulsating star icon button next to the timer. The button pulses for ~10 seconds, then stays visible until a rating is saved. Rating remains accessible from the triple-dot menu. ### Fixed - Fixed provider state mutation during widget tree build phase in timer session-ended detection by deferring to post-frame callbacks. - Fixed RenderFlex overflow on the rate button row by replacing the outlined text button with a compact star icon button. - Fixed race condition in realtime transcription stop: subscribe to the done-event stream before cleanup to avoid missing events on broadcast streams. ### Removed - Label assignment snackbar/toaster — no longer needed since label assignments are now confirmed individually via change sets. - Automatic rating modal trigger on timer stop — replaced by the pulsating rate button (see Changed above). ## [0.9.917] - 2026-03-12 ### Added - Task filter: filter by agent assignment (All / Has Agent / No Agent) in the Task Filter Modal. Enables efficient backfilling of agent assignments by showing only unassigned tasks. Zero performance overhead when filter is off. ### Fixed - Sync backfill: exact `(hostId, counter)` hits are now validated against the payload's current vector clock before resend. Stale exact mappings no longer resend a payload and then declare the same counter unresolvable in the same handling pass. ### Improved - Sync catch-up: when the stored marker cannot be found, catch-up now falls back to a bounded recent tail instead of replaying the entire visible room snapshot. This reduces repeated historical reprocessing during recovery. ## [0.9.915] - 2026-03-11 ### Removed - Legacy checklist updates prompt and task summary system — all conversation- based processing code, UI widgets, and associated tests have been removed. Task summaries are now handled exclusively by the agent system. - SQLite logging database (`LoggingDb`) — replaced by the existing file-based logging in `LoggingService`. Log files are written to `{docs}/logs/lotti-YYYY-MM-DD.log`, easily viewable with Console.app or any text editor. The in-app log viewer page has been removed in favor of native log viewers. ### Improved - Task summary resolution: a new `TaskSummaryResolver` provides unified fallback logic (agent report → legacy AI response) shared across prompt building, linked task context, and the chat summary tool. - Agent report selection: when multiple agent-task links exist, the resolver now picks the most recently assigned agent instead of an arbitrary one. - Recorder flow no longer blocks on transcription inference after recording. - Category settings no longer offer legacy prompt types that can never run. ## [0.9.914] - 2026-03-10 ### Fixed - Local inference model display: conversation headers now show the actual model name (e.g. qwen3.5:9b) instead of the stale template default (gemini-3-flash-preview), even while a conversation is still running. Resolved by reading the agent instance's inference profile directly. - Tool call count: deferred tools (set_task_title, update_task_priority, etc.) and locally-handled tools (update_report, record_observations) now correctly create action messages, so the conversation log shows accurate tool call counts instead of "0 tool calls". - Ollama JSON parsing: added resilient argument parser that handles markdown code fences, trailing text, and brace extraction from malformed JSON that smaller local models sometimes produce. - Ollama token usage: extract prompt_eval_count and eval_count from Ollama's final streaming chunk so token usage is recorded for local inference. ### Improved - Local model tool loop prevention: non-batch deferred tools can now only be called once per wake. Repeat calls are rejected with an explicit error, preventing smaller models from burning all turns on the same tool. - Guided tool responses: after a successful deferred tool call, the response lists remaining available tools, helping local models progress through title, priority, estimate, due date, etc. instead of repeating one tool. - Raised default maxTurnsPerWake from 5 to 10, giving agents more room to call multiple different tools in a single wake cycle. - Within-wake dedup for change set proposals: identical non-batch proposals are caught and skipped even before cross-wake dedup runs. - Ollama logging: full response chunks are now logged for easier debugging. ## [0.9.913] - 2026-03-10 ### Added - 35 new category icons: cycling, hiking, camping, pets, gardening, cooking, coffee, email, chat, video call, movies, podcast, theater, coding, crafts, dance, laundry, repair, banking, investment, receipt, celebration, gift, birthday, language, science, presentation, prayer, gratitude, self-care, stretching, weather, nature, volunteering, and recycling. ### Fixed - macOS crash on quit: replaced Dart `exit(0)` (which calls C `exit()` and triggers VM teardown/GC finalizers) with POSIX `_exit(0)` via FFI that terminates immediately. Also explicitly dispose media_kit Player before exiting so mpv's native core thread stops cleanly. Fixes SIGABRT from both mpv FFI callbacks and SQLite `NativeFinalizer` during VM teardown. - Change set UI stability: resolved tiles now match pending tile height (added tool name subtitle and matched icon size) to prevent vertical jumpiness when items are confirmed or rejected. - Label assignment: removed unnecessary rate limiter that blocked subsequent label confirmations in exploded change set batches. - Follow-up task handler: removed `sourceAudioId` parameter that the LLM could not reliably provide, eliminating UNIQUE constraint errors from hallucinated IDs. - Database: `upsertEntryLink` now guards against duplicate `(from_id, to_id, type)` links instead of crashing with UNIQUE constraint violation. ### Added - Profile-driven skills: introduced Skills as model-agnostic AI capability definitions (transcription, image analysis, etc.) and skill assignments on inference profiles to control which skills auto-trigger when assets are added to a task. Seven preconfigured skills are seeded on first launch. - Profile automation service: when an audio recording or image is added to a task whose agent's profile has the relevant skill with automation enabled, processing fires immediately via the profile's model slot — no legacy prompt lookup needed. Falls back to category-based automatic prompts when no profile-driven skill is configured. - Profile skill assignments UI: the inference profile form now shows an "Automated Skills" section where users can toggle automation per skill, with validation that the required model slot is populated. - Speech-recognition checkbox visibility: the opt-out checkbox now also appears when the task's agent profile has a transcription skill with automation enabled, not only when legacy category automatic prompts exist. ### Improved - Category icon picker: replaced flat gray styling with Material 3 semantic colors, subtle shadows, and selected-state glow for a more polished, dark-mode-friendly appearance. - Checklist dedup diagnostics: added logging of resolved title count in the change set builder for easier troubleshooting. ## [0.9.912] - 2026-03-09 ### Added - Qwen 3.5 local inference: added Qwen 3.5 9B and 27B models for local inference via Ollama with native multimodal, reasoning, and tool calling capabilities. - New "Local Power (Ollama)" inference profile for higher-end devices using Qwen 3.5 27B for thinking and Gemma 3 12B for image recognition. ### Changed - Updated "Local (Ollama)" inference profile to use Qwen 3.5 9B instead of Qwen3 8B. - Removed superseded local models: DeepSeek R1 8B, GPT-OSS 20B/120B, Qwen3 8B, and Gemma 3 12B QAT. - Default inference profiles now auto-update when upstream model IDs change, ensuring users always get the latest model assignments. - New known models are automatically backfilled for existing inference providers on app startup. ### Fixed - Ollama streaming tool call index: multiple tool calls streamed in separate chunks were merged into one malformed call because the loop index was used instead of the response's actual tool call index. - Flicker on task details page. ## [0.9.911] - 2026-03-09 ### Fixed - macOS crash on close: skip SQLite database close entirely on macOS before calling `exit(0)`. The previous fix still crashed because the SIGABRT occurred during `db.close()` itself (`sqlite3_close_v2` → `functionDestroy` → `DLRT_GetFfiCallbackMetadata`). SQLite WAL mode guarantees data integrity without explicit close. ## [0.9.910] - 2026-03-08 ### Added - Task agent follow-up task creation with deferred confirmation: the agent can propose splitting a task by creating a follow-up task and migrating checklist items, all presented for user review before execution. - Structured domain logging (DomainLogger) in the change set confirmation service, follow-up task handler, and checklist migration handler for better production diagnostics. ### Fixed - Checklist migration: LLM-hallucinated placeholder IDs are now replaced with the real deterministic placeholder, fixing "Target task lookup failed" errors when confirming migration items. - Checklist migration: copy is now created before archiving the source item, preventing data loss if the copy fails. - Checklist migration: archival failure no longer blocks the entire migration — returns success with a warning instead. - Follow-up task rejection now cascade-rejects sibling migration items that depend on the rejected task. - Follow-up task handler validates priority and dueDate args, rejecting malformed values. - Follow-up task handler checks createLink bool return value, surfacing link failure warnings. - Change set UI now shows partial-failure warnings in snackbars instead of silently discarding them. ## [0.9.909] - 2026-03-08 ### Added - Vector search distance telemetry: search result cards now show a color-coded cosine distance badge during vector search mode, and distance distribution (min/median/max) is logged for threshold calibration. ## [0.9.908] - 2026-03-07 ### Fixed - Sync backfill: agent entities and links are now included in automatic sequence log population at startup (previously only journal entries and entry links were populated, leaving agent counters unrecorded). - Sync backfill: added self-request guard to prevent hot-loop when echoed backfill requests arrive back from the Matrix room. - Sync backfill: "Reset Unresolvable" action in sync diagnostics allows recovery of entries incorrectly marked as permanently unresolvable. ### Changed - Sync performance: host activity cache (5-minute TTL) reduces redundant DB queries during incoming entry processing. - Localization: added ICU plural forms for reset success messages across all supported languages. ## [0.9.907] - 2026-03-07 ### Added - Individually rejectable label suggestions: agent-proposed labels are now shown as separate confirmable/rejectable items instead of a single bundle. - Rejecting a label automatically suppresses it so the agent does not re-propose it in future wakes. ## [0.9.906] - 2026-03-07 ### Added - Added an ObjectBox-backed embedding store POC for ANN search. ### Changed - Switched the current vector backend toggle to the ObjectBox path for macOS benchmarking, including the required sandboxed macOS app-group configuration. ## [0.9.905] - 2026-03-06 ### Added - Unified embedding generation: select multiple categories at once instead of one at a time. Includes Select All / Unselect All toggle. - Vector search now available on the journal tab (previously tasks-only). ### Changed - Refactored the embedding pipeline behind a backend-neutral store interface to prepare alternate vector backends without another large call-site migration. ### Removed - Removed "Re-index All Embeddings" maintenance action (superseded by multi-category generate). ## [0.9.904] - 2026-03-06 ### Fixed - Embedding backfill failing with "input length exceeds the context length" for entries containing code, URLs, or long unpunctuated text. The chunk target was lowered from 384 to 256 estimated tokens to leave sufficient headroom for content where the word-based heuristic undercounts. ### Added - Sequence log backfill now includes agent entities and agent links alongside journal entries and entry links, closing sync gaps for agent data. ## [0.9.903] - 2026-03-06 ### Fixed - AI task agent no longer repeatedly proposes checklist items that already exist on the task, were previously confirmed, or were explicitly rejected. - Pending change proposals are now included in the agent's LLM context so it avoids re-proposing items that are already queued for user review. ## [0.9.902] - 2026-03-06 ### Added - One-time backfill of vector clocks on agent entities and links created before the clock-stamping fix. Stamps oldest-to-newest, persists, and enqueues each for cross-device sync via the Sync Maintenance modal. ## [0.9.901] - 2026-03-05 ### Fixed - Task agent provider now reacts to sync notifications, so agents created on one device appear in the task detail view on other devices without a restart. - Removed unused collapsible section wrapper from linked entries widget. ### Added - Date-range re-sync for agent entities and links. The Maintenance re-sync action now includes agent data within the selected time interval instead of requiring a full sync of all agent records. ## [0.9.899] - 2026-03-04 ### Fixed - Break backfill amplification loop: superseded outbox entries now respond with hints (carrying the payload ID) instead of "unresolvable" flags, preventing the requester from re-requesting the same counter indefinitely. - Agent entities and links now record their vector clock counters in the sync sequence log on both send and receive paths. Previously, agent payloads consumed counters but never logged them, creating permanent phantom gaps that triggered unresolvable backfill requests and contributed to outbox spikes. ## [0.9.898] - 2026-03-04 ### Added - Overlapping document chunking for vector search embeddings. Long transcripts and agent reports are now split into overlapping chunks so every part of a long recording is semantically searchable, not just the first ~400 words. - Composite embedding index (entity_id + chunk_index) supporting multiple vectors per source document. - Re-index All Embeddings maintenance action to rebuild all embeddings with the new chunking strategy. ## [0.9.897] - 2026-03-03 ### Added - Enhanced agent observations with priority and category fields. Task agents can now flag observations as critical grievances or excellence notes, enabling targeted self-correction across wakes and prominent surfacing in 1-on-1 ritual sessions. - High-priority feedback card (GenUI): dedicated widget rendering grievances (red accent) and excellence notes (green accent) inline in evolution chat. - Ritual context builder injects a HIGH-PRIORITY FEEDBACK section before general feedback so improver agents address critical items first. - Task agents review their own prior critical observations at the start of each wake for self-correction. ## [0.9.896] - 2026-03-03 ### Added - Outbox priority queue: user-initiated actions (journal entries, entry links) now sync before bulk operations (entity definitions, tags, AI config). Three priority levels (high/normal/low) ensure responsive UX during large resyncs. - Sync health reporter: periodic health summary (every 5 min) logs outbox and sequence log counters when sync domain logging is enabled. - Domain-filtered sync logging in outbox send path via DomainLogger. ## [0.9.895] - 2026-03-03 ### Changed - Simplified audio recording modal: removed Checklist Updates and Task Summary checkboxes. These prompts now always use category defaults when configured, with no per-recording user toggle. - New system FTUE setup no longer populates automatic checklist update and task summary prompts on categories. Existing categories are unaffected. - Removed automatic task summary infrastructure (scheduled refreshes, smart triggers, countdown UI). Task summaries are now generated on demand by agents. Existing summaries remain viewable and deletable. ## [0.9.894] - 2026-03-02 ### Added - Vector search testing UI: toggle between full-text (FTS5) and vector (Ollama embedding) search on the tasks page. Shows timing and result count for evaluating search quality. Controlled by a new feature flag in Settings > Flags. ### Improved - Agent template settings tab UX: the 1-on-1 review button is now prominently placed in the sticky bottom bar when no edits have been made. Save/Cancel buttons only appear when the form has unsaved changes. - Directive text areas expand to full height instead of capping at 12 lines. - Profile selector label changed from "Inference Profiles" to "Default inference profile" for clarity. - Delete action moved from bottom bar to Stats tab to reduce clutter. ### Fixed - Evolution 1-on-1 review flow: restored the chat entry FAB so the interactive conversation can always be opened from the review page. ## [0.9.893] - 2026-03-02 ### Added - Voice transcription in evolution chat: mic button with batch and realtime recording modes, matching the existing AI chat transcription flow. Transcript populates the text field for editing before sending. ## [0.9.892] - 2026-03-02 ### Added - Split agent template directives into separate general directive (persona, tools, objectives) and report directive (output structure, formatting) fields. Existing templates are automatically seeded with purpose-built defaults. - Agent reports now support a structured TLDR field alongside the full report content. The TLDR is always visible with the full report expandable below. - Template edit page shows two distinct text areas for general and report directives instead of a single combined field. - Evolution proposals now present both directive fields independently with separate diff views. ## [0.9.890] - 2026-03-01 ### Added - Meta-improver agent: recursive self-improvement layer that evaluates and improves the template-improver agents themselves. The meta-improver runs monthly rituals focused on improver effectiveness, directive churn stability, proposal acceptance rates, and session outcome trends. - Seeded meta-improver template with meta-level directives for evaluating the improvement process rather than task-level performance. - Evolution session feedback extraction: the feedback pipeline now includes signals from evolution session outcomes (ratings, completion, abandonment) and directive churn detection for improver templates. - Recursion depth policy enforcement: improver agent creation validates against the maximum recursion depth (capped at 2 per ADR 0012). ## [0.9.889] - 2026-03-01 ### Fixed - Agent no longer proposes redundant changes that match the current state. Checklist items already checked/unchecked, estimates, priorities, due dates, statuses, and titles are now filtered at proposal time instead of cluttering the confirmation UI. The agent receives corrective feedback so it adjusts its reasoning. - Cross-device sync of confirmed/rejected change sets now works reliably. Agent entity payloads are fetched from the attachment descriptor before falling back to disk, preventing stale reads when the file download hasn't completed. Agent links are kept inline in text events for immediate availability. ## [0.9.888] - 2026-02-28 ### Added - Inference Profiles: named configuration bundles that group model assignments per capability slot (thinking, image recognition, transcription, image generation). Profiles replace direct model references on agent templates and instances, enabling intent-based selection like "keep data local" or "use EU infrastructure". - Profile management UI accessible from AI settings, with CRUD operations, deletion guards, and six seeded default profiles. - Two-page agent creation modal (template + profile selection) replaces the former single-step template picker. - Profile selector on both template detail and agent detail pages. ## [0.9.887] - 2026-02-28 ### Added - Agent now learns from past user decisions: confirmed, rejected, and deferred proposals are fed back into the agent's wake context so it avoids repeating rejected suggestions and builds on confirmed preferences. ### Fixed - Agent no longer proposes redundant label assignments when a task already has 3 or more labels. The available labels context is now omitted for fully-labeled tasks, preventing unnecessary change set proposals in the confirmation UI. ## [0.9.885] - 2026-02-28 ### Changed - Agent sync now uses file-attachment pattern (same as journal entities), preventing failures when agent payloads exceed Matrix's 60KB text event limit. Existing inline messages remain backward-compatible. ## [0.9.884] - 2026-02-28 ### Added - Domain-specific logging infrastructure with PII-safe sanitization and per-domain toggles (agent runtime, agent workflow, sync) in Settings > Advanced > Logging Domains. - Defensive fixes for agent wake orchestrator: synchronous timer scheduling before async DB write, drain timeout guard to prevent stuck locks, and error-level logging for wake run insert failures. ### Changed - Agent settings navigation now uses Beamer-native back navigation, fixing double-back-press and bottom-transition bugs in the Settings > Agents flow. - Agent report section on task detail: refresh icon replaces play icon when idle, play-now + countdown pill + cancel button during countdown, stable robot icon on the left with no layout jumps. - Responsive token usage table: vertical card layout on narrow screens (<600 px), standard table on wide screens. - Template detail page hides bottom bar on non-editable tabs (Stats, Reports). - Shortened "Save as New Version" button labels across all locales to prevent truncation on narrow screens. - Agent report section positioned above legacy AI summary in task form. ## [0.9.883] - 2026-02-28 ### Added - Checklist user sovereignty: track who last toggled each checklist item (user or agent) and prevent the AI agent from overriding user-set states without a substantive reason citing post-dated evidence. ## [0.9.882] - 2026-02-28 ### Added - Aggregate token usage tracking per template with tabbed detail page (Settings / Stats / Reports) showing per-model summaries across all instances and per-instance breakdowns with lifecycle badges. ## [0.9.881] - 2026-02-28 ### Added - Token usage section on agent detail page showing aggregated input, output, thinking, and cached token counts per model with grand totals. ## [0.9.880] - 2026-02-27 ### Added - Bulk-delete button for old AI task summaries, with confirmation dialog showing the count of summaries to be deleted. ## [0.9.879] - 2026-02-27 ### Changed - Task-agent linked task context now uses latest linked task-agent reports instead of legacy linked task summaries. - Agent state refresh for persisted throttle updates now routes via `UpdateNotifications` (removing the dedicated orchestrator state stream). - Template MTTR chart data loading now de-duplicates linked task lookups and resolves agent/task queries concurrently. ## [0.9.878] - 2026-02-27 ### Added - Settings > Agents page with Templates and Instances tabs, replacing the standalone Agent Templates page. - Agent instances list with kind filter (All / Task Agent / Evolution) and lifecycle filter (All / Active / Dormant / Destroyed). - Expandable TLDR agent report section on task detail page, shown when an agent has produced a report. - Agent report format aligned with task summary structure (TLDR, Achieved, Remaining, Learnings). ### Changed - Agent default model changed from Gemini 3.1 Pro to Gemini 3 Flash. - Agent report section uses expandable TLDR pattern instead of plain markdown. - Agent report tool description now references system prompt structure and encourages agents to express their personality. ## [0.9.877] - 2026-02-27 ### Added - Alibaba Cloud (Qwen) as AI inference provider with 5 models including text, vision, reasoning, audio, and image generation variants via DashScope API. - Qwen3 Omni Flash for audio transcription via Alibaba provider. ## [0.9.876] - 2026-02-27 ### Added - Outbox payload size tracking: records total payload size (JSON + file attachment) for each outbox item and displays it in the outbox monitor list. - Daily outbox volume aggregation query with configurable time window. - Database migration (sync DB schema v5) adding `payload_size` column to outbox table. - Daily sync volume bar chart on the Outbox Monitor page showing 30 days of outbox throughput in KB. - `headerSliver` slot on `SyncListScaffold` for inserting content between the filter header and the list. ## [0.9.875] - 2026-02-27 ### Fixed - Task agent automatic trigger broken after throttle optimization: fixed `_scheduleDeferredDrain` silently dropping drain when deadline already expired, added periodic safety-net timer to recover stuck queue jobs, and fixed startup hydration race that could cancel just-scheduled deferred drain timers. - Agent template model selection not persisted when editing: changing the AI model in edit mode now correctly saves the updated model ID to the template entity. ## [0.9.874] - 2026-02-27 ### Added - Evolution chat: multi-turn GenUI conversation for reviewing agent template performance, providing structured feedback, and approving LLM-rewritten directives. - Evolution dashboard with 2x2 mini chart grid: success rate sparkline, wake history bar chart, version performance comparison, and MTTR trend. - MTTR chart now measures true task resolution time (agent creation to task DONE/REJECTED) instead of individual wake run execution duration. - GenUI catalog integration for evolution workflow: proposal cards with approve/reject actions, note confirmation cards with expand/collapse, and version comparison diffs. - Labels included in agent task summary reports. ### Fixed - `assign_task_labels` tool crash due to `LabelsRepository` not being available outside Riverpod context. - Countdown timer digit "breathing" fixed with tabular figures font feature. ## [0.9.873] - 2026-02-27 ### Added - Agent tools: `set_task_language`, `set_task_status`, and `assign_task_labels` for automated task metadata management. - Correction examples context injection: agent receives per-category transcription correction patterns for improved checklist editing. - Label context injection: agent receives assigned, suppressed, and available labels for informed label assignment. - Defer-first throttle: initial agent wake deferred by 120 seconds to coalesce bursty edits, down from 300-second post-execution cooldown. - Standardized agent report structure with TLDR, Achieved, What's left, and Learnings sections. - ASR re-trigger: audio entry updates now notify the parent task, enabling agent wakes when transcription arrives. ## [0.9.872] - 2026-02-26 ### Added - Agent throttle gate: 5-minute cooldown between automatic subscription-triggered agent wakes to reduce token consumption during rapid task mutations. - "Run Now" button on task agent chip to bypass throttle and trigger immediate agent execution. - Countdown timer on task agent chip showing remaining throttle cooldown time. - Throttle deadline persistence via `nextWakeAt` field, surviving app restarts and backgrounding. ## [0.9.871] - 2026-02-26 ### Fixed - Reference image selection not rendering in cover art generation modal due to unbounded height constraints from the modal's scrollable content area. ## [0.9.870] - 2026-02-24 ### Added - Agent templates: full-stack feature for creating, editing, and managing reusable agent templates. Includes database schema, repository, service layer, Riverpod providers, and UI (list page, detail page with model selector). - Template evolution workflow: LLM-assisted 1-on-1 page for reviewing agent performance metrics, providing structured feedback, and having the LLM rewrite template directives for approval. - Template performance metrics: track token usage, execution counts, and success rates per template. - Inference provider resolver for selecting AI models per template. - Settings navigation for agent templates (list, create, edit). ## [0.9.869] - 2026-02-23 ### Added - Provider filter for category prompt selection: when multiple AI providers exist, filter chips let users narrow the prompt list by provider. - Collapsible text entries in linked-entry context: text notes can now be collapsed/expanded like images and audio entries. - Agent cross-device sync: agent entities and links are now synchronized via Matrix using `SyncAgentEntity` and `SyncAgentLink` message variants. Incoming agents restore wake subscriptions so task agents resume monitoring on the receiving device. - Sync maintenance support for agents: the maintenance UI includes "Agent Entities" and "Agent Links" re-sync steps that broadcast all local agent data to peer devices. - Zone-based transaction isolation in `AgentSyncService`: outbox messages are buffered during transactions and flushed only on successful commit. Nested transactions are supported via a depth counter; rollback discards all buffered messages. ### Fixed - Desktop app no longer crashes on exit: `WindowService` now disposes `OutboxService` and `MatrixService` before destroying the window. - Reference image selection: the "Continue" button is no longer hidden when many images are available. The image grid now scrolls independently while the button stays pinned at the bottom. ## [0.9.868] - 2026-02-22 ### Added - Agent cross-device sync via Matrix: agent entities and links synchronize across devices with wake subscription restoration. - Sync maintenance UI supports agent re-sync steps. - Zone-based transaction isolation: outbox messages are buffered until commit, with nested transaction and rollback support. ## [0.9.867] - 2026-02-22 ### Added - Agent running-state feedback: reactive spinner indicators on the task page agent chip and agent detail page app bar show when an agent is actively executing. - Agent inspectability: tool call arguments and results are now persisted as message payloads and viewable in expandable activity log cards with monospace formatting. User messages sent to the LLM are also persisted. - Conversation view: agent detail page now has tabbed Activity/Conversations views, with the conversation tab grouping messages by wake cycle (thread). - Post-execution signal drain: signals arriving while an agent is running are no longer lost — a 30-second drain timer picks up deferred work after execution completes. ## [0.9.866] - 2026-02-21 ### Added - Task Agent: persistent AI agents that maintain task summary reports and perform incremental metadata updates via tool calls (estimates, due dates, priorities, checklist items, titles). Gated behind the `enableAgents` config flag with a dedicated `agent.sqlite` database, wake orchestration with self-notification suppression, fail-closed category enforcement, and an agent detail inspection page with activity log, report viewer, and lifecycle controls (pause/resume/destroy). ## [0.9.865] - 2026-02-18 ### Added - Real-time transcription via Mistral Voxtral WebSocket API with live subtitles during recording (~2s latency). Available in both AI chat and the journal audio recording modal via a mode toggle. - Native WAV-to-M4A audio conversion via platform channels on iOS and macOS. ## [0.9.864] - 2026-02-18 ### Improved - Checklist swipe backgrounds now extend edge-to-edge, filling the full card width when swiping to archive or delete. ### Added - Archived checklist items are now marked as such in AI prompt context, with clear instructions that they are not active work — handled elsewhere, no longer relevant, or kept for reference only. ## [0.9.863] - 2026-02-17 ### Added - Archive swipe gesture for checklist items: swipe right to archive, preserving historical context without marking items as completed. Archived items are excluded from completion metrics and hidden in "Open" filter mode. ## [0.9.861] - 2026-02-16 ### Changed - Upgraded Flutter to 3.41.1 (Dart 3.11.0). - Removed legacy `EXCLUDED_ARCHS[sdk=iphonesimulator*] = arm64` from iOS Podfile, fixing simulator builds on Xcode 26+. - Added `iphonesimulator` to `SUPPORTED_PLATFORMS` in Xcode project settings, restoring simulator destination discovery. ## [0.9.860] - 2026-02-13 ### Changed - Enabled SQLite WAL mode, busy_timeout (5s), and synchronous=NORMAL for all databases via setup callback. Prevents "Database is Locked" errors when multiple isolates access the same file. - Added read pool of 4 isolates to JournalDb for offloading heavy reads from the UI thread. - Added `background` and `readPool` parameters to `openDbConnection()` and forwarded directory providers to JournalDb, SettingsDb, and SyncDatabase constructors, preparing for actor-based sync isolate. ## [0.9.859] - 2026-02-13 ### Changed - Removed artificial 200ms delay from sync outbox message enqueueing, improving outbox throughput. - Reduced user-activity idle threshold from 3s to 1.2s so the outbox resumes sending sooner after user interaction. ## [0.9.858] - 2026-02-13 ### Changed - Replaced Drift table-watcher streams with notification-driven streams for categories, habits, dashboards, measurables, labels, tags, and settings. Eliminates redundant DB polling and prepares for isolate-based sync. - Deduplicated notification-driven stream helpers into a single generic implementation. ### Fixed - Sync deserialization: unknown enum values from newer devices (e.g., new AI provider types) no longer cause infinite retry loops; unrecoverable `ArgumentError`/`FormatException` errors are logged and skipped. - Config flag init order: `initConfigFlags` now runs before `EntitiesCacheService.init()` so the private-entries flag reads its seeded default (`true`) on fresh installs instead of `false`. - Tags late-subscriber replay: `TagsService.watchTags()` now emits cached tags immediately for late subscribers instead of waiting for the next notification. - Theming sync race: `ThemingController` subscribes to setting notifications before the initial async load, preventing sync theme updates from being dropped during the init window. ## [0.9.857] - 2026-02-12 ### Changed - Removed M4A-to-WAV audio conversion for Mistral — all providers now accept M4A natively. Eliminates FFmpeg dependency and file size quota issues. - Updated Voxtral model description to reflect Voxtral Transcribe 2 capabilities with support for 13 languages and up to 1 GB / 3 hours per file. ## [0.9.856] - 2026-02-11 ### Fixed - Emoji Verification Flow: SAS verification now reacts to SDK `onUpdate`/`isDone` state changes, triggers a post-verification device-key refresh, and forces a sync rescan so sync resumes without restarting desktop or mobile. - Verification UX: Both outgoing and incoming verification modals now provide clearer "waiting for other device" feedback during confirmation and transition more reliably to the success state. - Verification Status Refresh: After SAS verification completes, the app now retries unverified-device refresh for a short window so the stale "unverified" badge clears without manual refresh, and desktop provisioning auto-advances from the QR step to the final status page once trust state is updated. - Provisioned Sync Disconnect: Disconnect now uses an explicit confirmation dialog (`Cancel` + destructive confirm) and closes the status modal after successful deletion. - Logging Performance: Sync-related logging no longer forces a disk flush for every non-test log line. File writes are now buffered and batched, with forced flush reserved for error-level paths. ### Changed - Sync Onboarding: The legacy manual Matrix setup wizard was removed from Settings, leaving provisioning as the single onboarding path while keeping emoji/SAS verification available from the provisioned flow. ## [0.9.855] - 2026-02-10 ### Added - Matrix Provisioning Wizard: New sync onboarding flow that lets a desktop device paste or scan a Base64-encoded provisioning bundle (homeserver, user, password, room ID), log in, join the sync room, rotate the password, and display a QR code with updated credentials for a mobile device to scan and complete setup. Includes full validation, step-by-step progress UI, error handling with retry, and a status page showing the connected account. ### Changed - Flexible Ratings Model: Session ratings are now driven by a catalog registry instead of hardcoded questions. Each rating dimension snapshots its question metadata (question text, input type, option labels) at save time, making stored ratings fully self-describing. Unknown catalogs (e.g. received via sync from a newer client) render in read-only mode with a graceful fallback chain: stored metadata, catalog lookup, then dimension key. - Matrix Sync Room Creation: Sync rooms are now created with encryption state in the initial room creation request (instead of relying on a later room snapshot), and room names now include the creating username for easier server-side identification. ## [0.9.852] - 2026-02-09 ### Fixed - Calendar Navigation: Tapping a time recording in the calendar view could navigate to a session rating instead of the parent task. The calendar link resolver now filters out RatingLinks, so only the Task-to-TimeRecording relationship is used for navigation. - Session Rating Category: Newly created session ratings inherit the category of the time entry they rate, so they display the correct task icon instead of a generic one. - Timer Entry Category: Starting a timer from a task now forwards the task's category to the new time recording, ensuring it is explicitly stored rather than left blank. - DailyOS Migration: The old Calendar tab was removed and replaced by the DailyOS tab. The classic calendar day view was removed and `/calendar` now always opens DailyOS. The old calendar and DailyOS feature flags were replaced by a single `enable_daily_os_page` flag. ## [0.9.851] - 2026-02-09 ### Fixed - Task Time Summation: Overlapping linked entries (e.g., a gym trip containing a fitness workout) are no longer double-counted when calculating time spent on a task. The task progress view now uses the same union-of-time-ranges algorithm that already powered the accurate Time Budget view, so both display consistent durations. ## [0.9.850] - 2026-02-08 ### Added - Collapsible Linked Entries: Image and audio entries linked to a task can now be individually collapsed and expanded with a chevron toggle. When collapsed, a compact preview shows a thumbnail (for images) or mic icon with duration (for audio) alongside the date, so you can identify entries at a glance without scrolling past large media. Collapse state is saved per link and synced across devices. - Collapsible Section Header: The "Linked Entries" section header now has a collapse chevron to hide the entire section at once, with smooth animated expand/collapse transitions. ## [0.9.849] - 2026-02-08 ### Added - Session Ratings: After stopping a timer that ran for at least one minute, a quick-rating modal prompts you to rate the session across four dimensions — productivity, energy, focus, and challenge-skill balance. Ratings are stored as journal entities linked to the time entry, with full sync support. - Session Ratings Config Flag: A new "Enable Session Ratings" toggle in Settings controls whether the rating prompt appears. Disabled by default. ### Fixed - Linked Entries Deserialization: Adding a new EntryLink variant (RatingLink) broke deserialization of all pre-existing links in the database, causing linked entries to disappear entirely from task detail pages. Fixed by using a freezed fallback union. ## [0.9.848] - 2026-02-08 ### Added - Czech Language: Full Czech (čeština) localization with ~1360 translated strings covering the entire app UI. - Platform Locale Registration: Added missing locale entries for iOS, macOS, and Windows so that all supported languages (cs, es, fr) are properly declared on every platform. ## [0.9.847] - 2026-02-07 ### Fixed - Audio Transcription: Single-speaker recordings no longer include an unnecessary "Speaker 1:" prefix. Speaker labels now only appear when multiple distinct voices are detected. - Checklist Creation: AI-powered checklist generation is now snappy regardless of the model used. Multiple individual function calls from the model are coalesced into a single batch database write, eliminating incremental UI updates. - Checklist Prompt: Strengthened instructions to enforce single-call batch creation, preventing models like Gemini Flash from splitting items across multiple function calls. - Deprecated Handler: The legacy `add_checklist_item` function is no longer processed; models that hallucinate this function name receive a redirect to the batch API. - Image Analysis: URLs extracted from screenshots are now always returned as proper Markdown links. URLs lacking a protocol (e.g., browser address bars showing `github.com/pulls`) are automatically prefixed with `https://`. ## [0.9.846] - 2026-02-07 ### Fixed - Automatic Image Analysis: Images imported via the photo library/camera roll now trigger automatic AI image analysis when configured on the category. Previously only paste and drag & drop triggered analysis; photo library imports required manually opening each image. - Automatic Image Analysis: Screenshots now also trigger automatic AI image analysis when configured on the category. ## [0.9.845] - 2026-02-06 ### Added - Gamey Design System: A complete vibrant visual theme with playful gradients, glow effects, and animations - Gamey color palette with feature-specific colors (journal teal, habit green, task blue, mood lavender, AI cyan, etc.) - Gradient system with feature gradients, reward tiers (gold/silver/bronze), neon, shimmer, and background gradients - Glow effects system with subtle/medium/strong/intense intensity levels and feature-specific glows - Animation utilities with spring curves, durations, and staggered animation support - `GameyThemeBuilder.apply()` integrates the gamey theme without modifying existing FlexColorScheme infrastructure - Gamey Widgets: Full set of themed UI components - `GameyCard` and `GameySubtleCard` with gradient borders, glow effects, and tap animations - `GameyFab` with custom character image and press animation for the gamey theme - `GameyJournalCard` and `GameyTaskCard` for themed list items - `GameySettingsCard`, `GameyToggleCard`, and `GameySettingsSection` for settings pages - `GameyIconBadge` with gradient background and optional glow - `GameyProgressBar` with animated gradient fill and shimmer effect - `AdaptiveSettingsCard` that auto-selects gamey or standard styling based on active theme - Theme Selection: Theme picker dropdowns now include the Gamey theme option with localized labels ### Fixed - Theme Selection: Replaced `TextEditingController` with `InputDecorator` in `SelectTheme` for proper dropdown behavior - Card Wrapper: Fixed `getIt.reset()` usage in test teardown for reliable test isolation - Gamey FAB: Prevented double invocation of `onPressed` callback - Gamey Task Card: Localized task status labels and added null-safety for `coverArtId` - Gamey Settings Card: Honored `accentColor` parameter, passed `showChevron` and `iconGradient` correctly - Gamey Card Animation: Handled `animateOnChange=false` in `didUpdateWidget` to avoid unnecessary animations ## [0.9.844] - 2026-02-05 ### Added - Cover Art Reference Images: Select up to 3 images linked to a task to guide the AI's visual style when generating cover art - New reference image selection step appears before cover art generation - Images are processed (resized to max 2000px, compressed to JPEG) before sending to Gemini - AI prompt updated to incorporate reference image style, materials, and atmosphere - Skip option available if no reference images are desired - Auto-skips if the task has no linked images - Journal Repository: New `getLinkedImagesForTask()` method for retrieving images linked to a task - Cover Art Cleanup: Deleting an image that's used as task cover art now automatically clears the `coverArtId` reference ### Fixed - Image Generation Modal: Button overflow issues on smaller screens resolved with flexible layout - Compact Task Progress: `showTimeText` parameter now defaults to platform-appropriate behavior (desktop shows time, mobile hides) ## [0.9.843] - 2026-02-05 ### Changed - Daily OS Time Budgets: Optimized UI for tasks with no budgeted time - Scenario A (no budget, no time recorded): Shows compact "No time budgeted" badge instead of cluttered "0m / 0m" + "0m left" + warning banner - Scenario B (no budget, time recorded): Shows "Xm / 0m" with "No time budgeted" badge instead of confusing negative time remaining - All badges now consistently right-aligned across all budget cards - Bordered badge styling with warning icon for clear visibility ## [0.9.842] - 2026-02-04 ### Added - Daily OS Timeline: Drag-and-drop support for planned time blocks - Long-press a block to initiate drag mode with haptic feedback - Move mode: drag the block body to reschedule the entire block - Resize mode: drag top/bottom edges to adjust start/end times - Section-bounded: blocks stay within their visible timeline section - Visual feedback: elevated styling, time labels, and duration indicator during drag - Small blocks (under 48px) use move-only mode to avoid cramped resize handles - 5-minute snap grid for precise scheduling - Disabled for blocks overlapping compressed timeline regions (expand first) ## [0.9.839] - 2026-02-03 ### Fixed - Audio Transcription: Prevent speaker name hallucinations in multi-speaker recordings - AI prompts now explicitly require "Speaker 1:", "Speaker 2:", etc. labels - Prohibits guessing speaker identities from voice characteristics or context - Speech dictionary terms are now only used for spelling corrections, not speaker identification ## [0.9.838] - 2026-02-03 ### Changed - Daily OS Time History Header: Refined visual design for a cleaner, more polished look - Selected day indicator now compactly wraps weekday and day number instead of extending over chart - Weekend days also use compact indicator styling for visual consistency - Removed vertical dividers between days for a cleaner appearance - Repositioned chart to sit below day labels with better visual separation - Wider selection boxes for improved touch targets and visual balance ## [0.9.837] - 2026-02-02 ### Added - Daily OS Time Budgets: Quick task creation button on each budget row - Tap the "+" button to instantly create a task with the category pre-assigned - Due date automatically set to the selected day - Navigates directly to task detail screen for immediate editing - Reduces friction when planning tasks during daily review ## [0.9.836] - 2026-01-31 ### Added - Daily OS Time History Header: New horizontally scrollable day navigation header - Replaces arrow-based navigation with tap-to-select day segments - Sticky month label shows the currently visible month(s) - Selected day highlighted with primary color border - Infinite scroll loads more history (14 days at a time) - Today button appears when not viewing current day - Displays day label chip and budget status indicator - Daily OS Time History Header: Stream chart visualization showing time-by-category distribution across days with symmetric stacked areas using the graphic library ### Changed - Daily OS Time History Header: Avoid full stacked-height recomputation when loading more history unless the scale changes - Daily OS Time History Header: Scale the stream chart to the maximum daily total with a 6-hour minimum - Daily OS Time History Header: Removed 180-day history cap for continuous exploration; users can scroll through unbounded history while memory remains efficient through lazy loading and viewport-based chart rendering ## [0.9.835] - 2026-01-31 ### Changed - Daily OS Time Budgets: Tasks completed on a different day now show a faded checkmark (45% opacity) - When viewing a past date where a task was scheduled but completed later, shows faded checkmark instead of hollow circle - Tasks completed on the viewed day still show full-opacity checkmark - Removes italic text styling for completed tasks for cleaner appearance - Daily OS Time Budgets: P2 (Medium) priority badges now visible alongside other priorities - Previously only P0, P1, P3 badges were shown; P2 was hidden as the default ## [0.9.834] - 2026-01-31 ### Changed - Daily OS Time Budgets: Redesigned with slim two-row header for reduced vertical space - Row 1: Category name, task completion indicator (count + circular progress ring), expand/collapse toggle - Row 2: Time recorded/planned, fixed-width progress bar, status text - Progress bar uses consistent green/red coloring (not category colors) and aligns in columns - Simplified typography with fewer font size/weight variants - Daily OS Focus Mode: Time budget cards auto-collapse based on active time block - When inside a planned block, the matching category stays expanded while others collapse - Helps focus attention on the current activity without manual toggling - Manual expand/collapse still available for any category ### Fixed - Daily OS Focus Provider: Fixed Riverpod violation where ref.watch was called inside async loop - Moved dependency watches before async operations to comply with Riverpod rules - Stream properly recreates when selected date or unified data changes - Daily OS Timer Updates: Added error handling when refetching data after timer stops - Errors during refetch are now logged instead of being silently swallowed - Matches error handling pattern used elsewhere in the controller ## [0.9.833] - 2026-01-30 ### Added - Voice Task Priority: Set task priority via voice during audio recordings - Say "priority P1", "this is urgent", or "low priority" to set task priority - AI maps spoken terms to priority levels: urgent/critical→P0, high/important→P1, medium/normal→P2, low/minor→P3 - Only sets priority when not already explicitly set, preserving manual edits - Follows the same pattern as voice-controlled time estimates and due dates ## [0.9.832] - 2026-01-30 ### Added - Daily OS Timeline: Double-tap planned blocks to edit start/end times - More intuitive alternative to the existing long-press gesture - Opens the same edit modal for adjusting time range, category, and notes ## [0.9.831] - 2026-01-29 ### Added - Daily OS Smart Timeline Folding: Timeline now intelligently compresses inactive periods - Gaps larger than 4 hours between entry clusters are automatically folded - Compressed regions display at 8px per hour (vs 40px normal) with a zigzag edge pattern - Tap any compressed region to expand it to full height - Hours before 6 AM and after 10 PM are folded by default when no entries exist - ±1 hour buffer around each entry keeps context visible - Dramatically reduces vertical scrolling on days with sparse activity ## [0.9.830] - 2026-01-29 ### Fixed - Daily OS Time Calculation: Overlapping time entries within the same category no longer double-count - A 1.5h "Gym Trip" containing a 45m "Fitness Entry" now correctly reports 1.5h total (not 2.25h) - Uses time range union algorithm to calculate actual time coverage without overlap inflation - Applies to all budget progress calculations and time summaries ### Changed - Daily OS Timeline: Same-category entries that fully contain another now nest visually - Child entries (e.g., the workout) render inset inside their parent block (e.g., the gym trip) - Reduces visual clutter by not using separate lanes for related entries - Different-category entries still display in separate lanes when overlapping ## [0.9.829] - 2026-01-29 ### Added - Daily OS Priority Indicators: Tasks now display priority badges (P0, P1, P3) styled like Linear - Compact color-coded badges: red for P0 (Urgent), orange for P1 (High), gray for P3 (Low) - P2 (Medium) badges are hidden to reduce visual noise since it's the default priority - Priority badges appear in both list and grid views ### Changed - Daily OS Future Date Filtering: Viewing future dates now shows only tasks due on that specific day - Overdue tasks no longer pollute future day planning views - Today's view continues to show both due-today and overdue tasks for review/rescheduling - Daily OS Priority Sorting: Tasks without tracked time are now sorted by priority first - Sort order: time spent (desc) → priority (P0→P1→P2→P3) → urgency → alphabetical - Daily OS Timeline: Entry heights now reflect actual duration based on calendar scaling - Removed 20dp minimum height constraint from planned and actual blocks - Very short entries appear as thin colored lines proportional to their duration ## [0.9.828] - 2026-01-28 ### Added - Daily OS Due Task Visibility: Tasks due today or overdue now appear in time budget cards - Due tasks show even without tracked time, ensuring nothing falls through the cracks - Visual badges indicate urgency: red "Overdue" / "Late" and orange "Due today" / "Due" - Categories with due tasks but no planned budget show a warning banner - Intelligent deduplication: tasks with both tracked time and due dates merge properly - Smart sorting: tasks with time first (by duration), then due tasks, then alphabetical ### Changed - Task view mode preference (list/grid) now persists per category - Stored in settings database, survives app restarts - Default changed from grid to list view for better readability ## [0.9.827] - 2026-01-27 ### Added - Daily OS Time Budgets: Task visualization with time tracking - Shows tasks that contributed time to each budget category - Displays per-task time spent and completion status (checkmark for tasks completed today) - Switchable list/grid view with toggle button - Grid view shows task thumbnails (cover art) with time badge and title overlay - List view shows compact rows with status indicator, title, and time - Collapsible task section with header showing task count and total time - Tasks sorted by time spent (descending), zero-time completed tasks at end ## [0.9.826] - 2026-01-27 ### Added - Daily OS Timeline: Overlapping entries now display in separate lanes - Uses greedy lane assignment algorithm to prevent visual overlap - Entries that don't overlap share the same lane (full width) - Overlapping entries are placed in adjacent lanes with equal widths - Small gap between lanes for visual separation ## [0.9.825] - 2026-01-27 ### Changed - Daily OS Architecture: Unified data controller for consistent real-time updates - Consolidated day plan, timeline data, and budget progress into single atomic state - Uses `ref.keepAlive()` + manual `StreamSubscription` to prevent Riverpod 3's automatic pausing - All UI components (timeline, budget progress bars, summary) now update together - Fixes auto-update failure when time entries are created or synced ### Removed - Deprecated controllers: `TimelineDataController` and `TimeBudgetProgressController` classes - Data types (`TimelineSlot`, `TimeBudgetProgress`, etc.) preserved in existing files - All functionality now handled by `UnifiedDailyOsDataController` ## [0.9.824] - 2026-01-27 ### Fixed - Day Plan Timeline: Fixed crash when entries cross midnight (e.g., 23:00 to 01:00 next day) - Entries ending after midnight are now correctly treated as ending at 24:00 for display purposes - Added defensive bounds checking to prevent RangeError in List.generate - Resolves blank/grey screen issue on days with late-night or overnight entries ## [0.9.823] - 2026-01-24 (WIP) ### Added - Daily Operating System: New time management feature for planning and tracking your day (WIP) ## [0.9.822] - 2026-01-24 ### Fixed - Task Time Tracking: Audio recording duration no longer counts toward time spent - Prevents double-counting when recording meetings (e.g., 1-hour recording + 1-hour time entry) - Only actual logged work entries contribute to task progress calculations ### Changed - Added missing translations to German, Spanish, French, and Romanian localization files ## [0.9.821] - 2026-01-24 ### Added - OpenAI Onboarding FTUE: First-time user experience for OpenAI providers with one-click setup - Auto-creates 4 models: GPT-5.2 (reasoning), GPT-5 Nano (flash), GPT-4o Transcribe (audio), GPT Image 1.5 (image) - Configures 9 optimized prompts with appropriate model assignments - Creates a test category with all prompts enabled and auto-selection configured - Follows the same streamlined pattern as Gemini FTUE ### Changed - Audio format handling simplified: M4A files now sent as-is labeled as mp3 for most providers - Only Mistral requires actual WAV conversion - Reduces processing overhead for OpenAI and Gemini transcription ## [0.9.820] - 2026-01-23 ### Changed - Upgraded dependencies ## [0.9.819] - 2026-01-23 ### Changed - Code Modularity Refactor: Extracted focused services from large modules for better testability and maintainability - `MetadataService`: Handles metadata creation with vector clock integration (from PersistenceLogic) - `GeolocationService`: Manages geolocation addition with race condition prevention (from PersistenceLogic) - `ExifDataExtractor`: Pure functions for EXIF timestamp and GPS coordinate extraction (from image_import) - `AudioMetadataExtractor`: Audio file metadata parsing and duration extraction via MediaKit (from image_import) - Media import separation: `audio_import.dart`, `media_import.dart`, and `image_import.dart` now have single responsibilities ## [0.9.818] - 2026-01-19 ### Changed - Voxtral Token Streaming: Upgraded from chunk-level to true token-by-token streaming - Text now appears progressively as it's generated, not waiting for full chunks - Tokens batched in groups of 6 for smooth display without excessive overhead - Configurable via Settings → Flags → "Enable AI streaming responses" - Non-streaming mode available for ~20% faster processing when progress display isn't needed ## [0.9.817] - 2026-01-18 ### Added - Mistral Cloud Provider: New cloud transcription option using Mistral's Voxtral API - Supports the `voxtral-small-2507` model for high-accuracy cloud transcription - Automatic audio format conversion from M4A to WAV (8kHz mono) for API compatibility - Uses FFmpegKit on iOS/Android/macOS, system FFmpeg on Linux/Windows - Auto-creates the Voxtral model when adding a Mistral provider ### Changed - Audio Transcription Prompts: Improved context handling to prevent transcript bleed - Now uses task summary instead of full task JSON for cleaner context - Added explicit instructions that context is for terminology only, not to be included in output - Transcriptions now include paragraph breaks on small pauses or topic changes ## [0.9.816] - 2026-01-18 ### Changed - Voxtral Streaming Transcription: Audio transcription now streams progressively as each 60-second chunk completes - Each chunk's transcription is sent as an SSE event, providing real-time visual feedback - No more waiting for the entire audio to process before seeing results - Particularly useful for longer recordings (2+ minutes) where chunk-level progress is meaningful - AI Streaming Enabled by Default: The `enableAiStreamingFlag` now defaults to `true` for new installations - Streaming responses provide better visual feedback during AI inference - Existing users can enable in Settings → Flags → "Enable AI streaming responses" ## [0.9.815] - 2026-01-16 ### Added - Voxtral Local Transcription: New local AI transcription service using Mistral's Voxtral model - Supports up to 30 minutes of audio transcription (vs 5 minutes for Gemma 3N) - 9 languages with automatic detection (English, Spanish, French, Portuguese, Hindi, German, Dutch, Italian, Arabic) - Context-aware transcription with speech dictionary support for improved accuracy - Two model options: Voxtral Mini 3B (~9.5GB VRAM) and Voxtral Small 24B (~55GB VRAM) - Runs locally on Apple Silicon (MPS) or NVIDIA GPUs (CUDA) - OpenAI-compatible API with full Flutter integration - New Python service at `services/voxtral-local/` on port 11344 ## [0.9.814] - 2026-01-12 ### Fixed - Gemini Setup Dialog Reset: Fixed reset not working properly - Migrated storage from SharedPreferences to SettingsDb (SQLite) for reliability - Reset now takes effect immediately without requiring app restart - Maintenance page now properly invalidates the provider state ### Changed - Localized the Gemini setup dialog reset strings in maintenance settings ## [0.9.813] - 2026-01-12 ### Changed - What's New Modal: Added "Done" button on the last page for clearer completion - "Skip" button hidden on last page, replaced with "Done" on the right - Button labels now use localization (l10n) for proper internationalization - Cover Art Display: Reverted to `BoxFit.cover` for better mobile display ### Fixed - Gemini Setup Dialogs: Fixed dialogs disappearing when resizing the app window - Added `useRootNavigator: true` to ensure dialogs survive widget tree rebuilds - Affects both the Gemini setup prompt modal and FTUE setup dialog ## [0.9.812] - 2026-01-12 ### Fixed - Task Estimate Field: Fixed estimate field not appearing on Linux (and fresh installs) - The estimate chip was incorrectly hidden when no labels were defined in the app - Now the estimate and "Add Label" controls always appear, regardless of label state ## [0.9.811] - 2026-01-11 ### Changed - Gemini FTUE Streamlined: Reduced default prompts from 18 to 9 with optimized model assignments - Gemini Pro: Checklists, Coding Prompts, Image Prompts (complex reasoning tasks) - Gemini Flash: Audio Transcription, Task Summary, Image Analysis (fast processing) - Nano Banana Pro: Cover Art generation (with reasoning mode enabled for reliable image output) - Gemini Setup Modal: Now dismissible by tapping outside (temporary skip) - Tapping outside closes the modal but doesn't permanently dismiss it - Modal will reappear on next app start, allowing users to set up later - "Don't Show Again" button still permanently dismisses the prompt ## [0.9.810] - 2026-01-11 ### Fixed - Flatpak: Fixed keyring access error by adding D-Bus permission for org.freedesktop.secrets - Flatpak: Fixed cursor theme warning by exposing host cursor theme paths via XCURSOR_PATH - Flatpak: Fixed fontconfig error by removing invalid FONTCONFIG_PATH/FONTCONFIG_FILE overrides from wrapper script - Flatpak: Fixed application icon warning by adding Flatpak-specific icon path to the icon search paths ## [0.9.809] - 2026-01-11 ### Fixed - Cover Art Display: Fixed image cropping on mobile devices where sides were cut off - Changed from BoxFit.cover to BoxFit.fitWidth with top alignment - Ensures full 16:9 cover art is visible without side cropping - Cover Art Set/Unset: Fixed cover art changes not persisting after database sync - Vector clock counter could fall behind synced data, causing updates to be rejected - Added auto-catch-up logic to detect stale counters and recover automatically ## [0.9.808] - 2026-01-11 ### Added - Voice-Controlled Task Properties: Set time estimates and due dates via voice during audio recordings - `update_task_estimate`: Converts natural language durations ("2 hours", "half a day") to task estimates - `update_task_due_date`: Resolves relative dates ("Friday", "next week") to absolute due dates - Current date injected into AI prompt for accurate relative date resolution - Only sets values when not already set (preserves manual edits) - Treats zero-duration estimates as "not set" for intuitive behavior ## [0.9.807] - 2026-01-10 ### Added - Gemini Quick Start: Automatic setup prompt for new users without AI providers configured - Modal appears after What's New is dismissed (no overlapping dialogs) - One-click Gemini setup with pre-configured models and prompts - Explains available AI features: audio transcription, image analysis, smart checklists, task summaries - "Don't Show Again" option persists dismissal permanently - Notes that other providers like Ollama for local inference are available in Settings > AI ## [0.9.806] - 2026-01-10 ### Changed - Upgraded to Flutter 3.38.6 ## [0.9.805] - 2026-01-09 ### Added - What's New Modal: Editorial magazine-style release notes viewer - 21:9 hero banner images with glassmorphism version badge - Multi-release navigation with animated indicator dots - Version filtering: only shows releases matching or older than installed app version - Smart "seen" tracking: only marks releases you actually viewed as seen - Image precaching for smooth page transitions (banners + markdown images) - "Skip" button to dismiss all remaining releases at once - "View past releases" option in empty state to review previous updates - Responsive design: bottom sheet on mobile, centered dialog on desktop - Markdown content support with images resolved from remote content repository ## [0.9.804] - 2026-01-08 ### Added - Linked Tasks Section: New dedicated UI section for viewing and managing task-to-task links - Displays bidirectional links with directional indicators (↳ LINKED FROM, ↗ LINKED TO) - Minimal, Linear-style text links with status circles (open/completed) - "Link existing task..." modal with searchable task list (FTS5-powered) - "Create new linked task..." creates a subtask inheriting current task's category - "Manage links..." mode shows unlink buttons for each linked task - Tasks filtered out from generic "Linked Entries" section to avoid duplication - Location: Between AI Task Summary and Checklists sections in task details ## [0.9.803] - 2026-01-08 ### Fixed - Calendar Auto-Refresh: Fixed calendar not updating when returning from creating a time recording - Improved `onVisibilityChanged` to track visibility transitions (invisible → visible) - Uses `ref.invalidateSelf()` on transition instead of direct fetch to ensure full provider rebuild - Added missing `ref.mounted` guard to `DayViewController.onVisibilityChanged` - More efficient than the removed `onResume` approach (no CPU overhead when staying on page) ## [0.9.802] - 2026-01-07 ### Changed - Scroll Performance Optimizations: Significant improvements to reduce jank during scrolling - Added `cacheExtent` to key scrollable views (Entry Details: 4000px, Tasks List: 1500px) - Labels Modal converted to sliver-based architecture with `SliverList.builder` for lazy loading - Added `RepaintBoundary` isolation for task cards, linked entries, and checklist items - Fixed memory leaks in EditorWidget and TextViewerWidget (ScrollController/FocusNode now properly managed) - These optimizations target smooth 60/120fps scrolling, especially for tasks with many checklist items ### Fixed - High CPU Usage on Calendar Page (#2578): Removed `ref.onResume(() => ref.invalidateSelf())` from calendar controllers - The pattern caused an infinite rebuild loop (160%+ CPU when idle) - Root cause: `invalidateSelf` on resume triggered rebuild → pause → resume → invalidateSelf cycle - Calendar updates are already handled by existing `onVisibilityChanged()` and stream listeners ## [0.9.801] - 2026-01-06 ### Changed - Entry Actions Menu Redesign: Restyled to match the FAB addition menu with Nano Banana Pro styling - Unified list design with subtle horizontal dividers (removed card-based layout) - Clean icons without gradient containers - Consistent with FAB menu visual language - Conditional item visibility handled at parent level to prevent double dividers ## [0.9.800] - 2026-01-06 ### Fixed - Calendar View Stale Data: Fixed calendar not updating after adding or modifying entries - Root cause: Riverpod 3's auto-pause feature pauses providers when widgets are not visible - Added `ref.onResume()` callbacks to `DayViewController` and `TimeByCategoryController` - Providers now invalidate themselves when resuming, ensuring fresh data is fetched - Uses `Future.microtask()` to defer invalidation per Riverpod 3 lifecycle callback restrictions ## [0.9.799] - 2026-01-05 ### Changed - FAB Addition Menu Redesign: Modernized the add entry bottom sheet with Nano Banana Pro styling - Unified container with single dark background (removed separate card backgrounds) - Clean list items with white icons (removed rounded gradient containers) - Subtle horizontal dividers between menu items - Consistent icon, text, and plus (+) trailing icon layout - Edge-to-edge tap effects for better visual feedback ### Fixed - OpenAI Compatible Provider: API Key field now visible when adding or editing providers (#2570) - Removed `genericOpenAi` from `noApiKeyRequired` set as most OpenAI-compatible endpoints require Bearer Token authentication - Authentication section now displays for OpenAI Compatible provider type ## [0.9.798] - 2026-01-04 ### Added - AI Cover Art Generation: Generate task cover art directly from audio descriptions using Gemini's image generation model (Nano Banana Pro) - Trigger from audio entry action menu when linked to a task - Full task context included in prompt (title, checklists, labels, linked task summaries) - `{{current_task_summary}}` placeholder provides learnings and annoyances for visual metaphors - Review modal with accept, edit prompt, and regenerate options - Accepted images automatically set as task cover art ### Changed - Cover Art Display: Updated from 2:1 to 16:9 aspect ratio for consistency with generated images - Task expandable app bar now uses 16:9 aspect ratio - Cover art background displays at full resolution (removed cacheHeight constraint) - Image Generation: Uses 2K resolution for Full HD quality output (1920x1080 at 16:9) ## [0.9.797] - 2026-01-04 ### Fixed - Cross-Checklist Drag-and-Drop: Restored ability to move items between checklists - Fixed regression from checklist redesign PR where `ReorderableDragStartListener` blocked `super_drag_and_drop` gestures - Moved `DropRegion` to wrap entire checklist card for proper drop target area - Added item-level `DropRegion` for position-aware drops enabling within-list reordering - Unified drag system: long-press anywhere on item initiates drag for both reordering and cross-checklist moves ## [0.9.796] - 2026-01-03 ### Changed - Nano Banana Checklist Redesign: Major UI rewrite with card-based architecture - Modular widget structure: extracted `ChecklistCardHeader`, `ChecklistCardBody`, `ChecklistFilterTabs` into separate files - Animated chevron rotation (90°) between expanded/collapsed states - Filter tabs with underline indicator (replaced SegmentedButton) - Progress ring hidden when empty (expanded mode), always shown (collapsed) - Add input moved to bottom of checklist body - Click-to-edit title functionality - Three display modes: expanded, collapsed, sorting - Comprehensive test coverage for all extracted components (32 new tests) ## [0.9.795] - 2026-01-03 ### Changed - Image Card Thumbnails: Now display as uniform squares with center crop - Portrait images cropped from vertical center - Landscape images cropped from horizontal center - Consistent 160x160 thumbnail size for visual uniformity - Updated Gemini AI models to latest versions ## [0.9.794] - 2026-01-02 ### Changed - Due Date Visibility Refinements: Improved UX for completed and rejected tasks - Due dates are now hidden on task cards for completed and rejected tasks - Due dates in entry details view show grayed-out styling for completed/rejected tasks (no red/orange urgency colors) - Updated README and Flatpak metainfo to reference the launched "Meet Lotti" blog series ### Added - Tests for due date visibility behavior based on task status ## [0.9.793] - 2025-12-31 ### Added - Visual Mnemonics: Tasks can now have cover art images for memorable visual representation - Set any linked image as a task's cover art via the image action menu - Cover art thumbnails displayed on task cards (toggle in filter modal) - Cinematic 2:1 collapsible SliverAppBar with cover art in task details - Glass effect icon containers for overlay buttons on cover art - Horizontal crop adjustment for cover art positioning ### Changed - Task detail app bar adapts between compact (no cover) and expandable (with cover) modes - Added comprehensive test coverage for cover art widgets ## [0.9.792] - 2025-12-31 ### Changed - Upgraded to Riverpod 3.x with comprehensive codebase migration - Updated all providers to use new Riverpod 3 syntax and patterns - Added `ref.mounted` checks after async gaps to prevent disposed ref access - Converted family providers to use constructor-based parameter passing - Updated notifier classes from `AutoDisposeFamilyAsyncNotifier` to `AsyncNotifier` pattern - Changed callbacks in widgets to use closures for fresh notifier instances at invocation time - Made `MatrixStats` class immutable with `@immutable` annotation and `final` fields - Added `ref.onDispose` cleanup handlers for resource management ### Fixed - Fixed "Cannot use Ref after disposed" errors in autoDispose providers - Fixed checklist item resolution to properly await async futures instead of reading sync values - Fixed provider override conflicts in tests by detecting duplicate overrides ## [0.9.790] - 2025-12-31 ### Added - Task Due Dates: Tasks now support due date assignment and display - Due date picker in task details header with Cancel/Clear/Done actions - Color-coded status indicators: red for overdue, orange for due today - Due date display on task cards with toggle in filter modal - Tappable due date text toggles between absolute ("Dec 24, 2025") and relative ("Due in 5 days") formats - Proper localization with ICU plural support for English, Spanish, and French - Shared `getDueDateStatus` utility for consistent status calculations - Uses `clock` package for testable time-dependent logic ### Changed - Task card layout: creation date on LEFT, due date on RIGHT - Task details header: improved styling with restyled filter modal - Added comprehensive test coverage for due date widgets and utilities ## [0.9.788] - 2025-12-30 ### Changed - Migrated journal/tasks page state management from Bloc to Riverpod - Replaced `JournalPageCubit` with `JournalPageController` notifier using `@Riverpod(keepAlive: true)` - Uses family provider pattern with `showTasks` boolean for journal vs tasks page state - Created `journalPageScopeProvider` for scoped access to `showTasks` value in widget tree - Updated `InfiniteJournalPage`, `JournalSliverAppBar`, filter widgets to use Riverpod - Fixed modal state sharing in `JournalFilterIcon` and `AiChatIcon` using `UncontrolledProviderScope` - Added comprehensive test coverage for controller, state, and widget interactions ## [0.9.787] - 2025-12-30 ### Fixed - Fixed habit settings page opening empty for existing habits - Changed `ref.watch()` to `ref.read()` for `habitByIdProvider` in controller build - Prevents controller rebuilds when stream data changes, preserving form state - Edit flow now loads cached habit data synchronously from `EditHabitPage` ## [0.9.786] - 2025-12-30 ### Changed - Migrated habits page state management from Bloc to Riverpod - Replaced `HabitsCubit` with `HabitsController` notifier using `@Riverpod(keepAlive: true)` - Created Freezed-based `HabitsState` with helper functions for chart calculations - Updated `HabitsTabPage`, `HabitsSliverAppBar`, `HabitStreaksCounter`, `HabitsFilter`, `HabitsSearchWidget`, and `HabitCompletionRateChart` to use Riverpod - Fixed TextEditingController lifecycle in search widget (proper init/dispose/sync) - Fixed chart touch handling to defer state modification via `addPostFrameCallback` - Added comprehensive test coverage for controller, state helpers, and widgets ## [0.9.785] - 2025-12-30 ### Changed - Migrated audio player state management from Bloc to Riverpod - Replaced `AudioPlayerCubit` with `AudioPlayerController` notifier - Uses `@Riverpod(keepAlive: true)` for app-wide audio state persistence - Updated `AudioPlayerWidget` to use Consumer pattern - Updated `RecorderController` to use ProviderContainer injection - Removed BlocProvider wrapper from app initialization ## [0.9.784] - 2025-12-30 ### Changed - Migrated habit settings state management from Bloc to Riverpod - Replaced `HabitSettingsCubit` with `HabitSettingsController` notifier - Uses family provider pattern with `habitId` as key for per-habit state management - Added `habitByIdProvider` for watching habit by ID from database - Added `habitDashboardsProvider` for dashboards in habit settings - Updated `HabitDetailsPage`, `CreateHabitPage`, `EditHabitPage` to use Riverpod - Updated category and dashboard selection widgets to use Riverpod providers ## [0.9.783] - 2025-12-30 ### Changed - Migrated theming state management from Bloc to Riverpod - Replaced `ThemingCubit` with `ThemingController` notifier and `enableTooltipsProvider` - Theme state now managed via `themingControllerProvider` with keepAlive for app-wide persistence - Moved theme definitions (`themes` map, `LightModeSurfaces`) to `lib/features/theming/model/` - Updated `MyBeamerApp` to use `ConsumerStatefulWidget` with Riverpod providers - Updated `ThemingPage` to use `ConsumerWidget` pattern ## [0.9.782] - 2025-12-30 ### Changed - Migrated sync outbox state management from Bloc to Riverpod - Replaced `OutboxCubit` with `outboxConnectionStateProvider` and `outboxPendingCountProvider` - Updated `OutboxBadgeIcon` to use Riverpod providers for consistency - Removed BlocProvider wrapper from app initialization - Moved `OutboxStatus` enum to new provider file ## [0.9.781] - 2025-12-29 ### Added - Task Sorting Options: Tasks can now be sorted by date (newest first) or by priority - New segmented button in task filter modal to toggle between sort modes - Priority sort (default): Orders by priority rank (P0→P1→P2→P3), then by date within same priority - Date sort: Orders by creation date descending, regardless of priority - Sort preference persists across sessions - Creation Date Display: Optional creation date shown on task cards - Toggle in filter modal to show/hide creation date on task list cards - Date displayed in bottom-right corner with subtle styling (small font, low contrast) - Uses locale-aware date format (e.g., "Dec 29, 2024") - Preference persists across sessions ### Changed - Improved accessibility for task filter toggle (now uses SwitchListTile) - Optimized date display toggle to not trigger unnecessary query refreshes ## [0.9.780] - 2025-12-29 ### Changed - Migrated dashboard list state management from Bloc to Riverpod - Replaced `DashboardsPageCubit` with Riverpod providers for consistency with codebase - Dashboard filtering and sorting now uses `filteredSortedDashboardsProvider` - Category filter selection uses `selectedCategoryIdsProvider` - Improved sorting to be case-insensitive for better UX ## [0.9.779] - 2025-12-29 ### Added - Labels for All Entry Types: Labels can now be assigned to any journal entry, not just tasks - Labels display on journal list cards (alphabetically sorted, privacy-filtered) - Labels display in entry detail views with edit capability - New "Labels" action in the triple-dot menu for quick label assignment - Unified label selection modal with search, create, and apply functionality - Consistent UX across all entry types (text, audio, image, event entries) ### Changed - Refactored label selection modal code into shared utility for better maintainability - Extracted `LabelSelectionModalUtils` for centralized modal handling - Fixed memory leak in label selector (proper disposal of controllers) - Apply button now correctly disables when no changes are pending ## [0.9.778] - 2025-12-28 ### Added - Twi Language Support: Added Twi (Akan) as a supported language for task summarization - Twi is spoken by approximately 9 million people, primarily in Ghana - Enables AI-generated task summaries in Twi - Includes Ghana flag in language selection UI ## [0.9.777] - 2025-12-28 ### Added - Image Prompt Generation: New AI response type for generating image prompts from task context - Transforms audio recording + task context into detailed prompts for AI image generators - Triggered from audio entries linked to tasks (same pattern as Coding Prompt) - Includes visual metaphor guidelines (debugging = bugs, progress = paths, etc.) - Style options: infographic, cartoon, artistic, photorealistic, retro, minimalist, isometric - Output format with Summary and Prompt sections for easy copy-paste - Designed for Midjourney, DALL-E 3, Stable Diffusion, Gemini Imagen - Uses GeneratedPromptCard UI with prominent copy button - Full localization support (EN, DE, ES, FR, RO) ## [0.9.776] - 2025-12-28 ### Added - Task Summary Goal Section: AI task summaries now include a succinct Goal section after the TLDR - Describes the desired outcome or essential purpose of the task (1-3 sentences) - Helps maintain focus on the "why" behind the work - Displayed in the expanded summary view alongside Achieved Results and Remaining Steps - Entry Labels Support: Labels can now be assigned to all entry types, not just tasks - Labels display on journal list cards (alphabetically sorted, privacy-filtered) - Labels display in entry detail views with edit capability - New "Labels" action in the triple-dot menu for quick label assignment - Consistent Wolt-style modal UX for label selection across all entry types ## [0.9.775] - 2025-12-27 ### Added - Single-User Multi-Device Sync: New room discovery feature for simplified sync setup - Device B can now discover and join existing sync rooms without invitation from Device A - Room Discovery page shows potential sync rooms with confidence indicators - Rooms are evaluated based on encryption status, Lotti state markers, and message content - Smart navigation skips discovery when a room is already configured - Bounded concurrency (5 parallel evaluations) for faster discovery with many rooms - Full localization support for room discovery (EN, DE, ES, FR, RO) ### Fixed - QR scanner now restarts properly after invite failure, allowing retry without navigation ## [0.9.774] - 2025-12-26 ### Added - Calendar Privacy Filtering: Calendar view now respects category visibility settings from Tasks page - Selected categories display colored box with text (title and description) - Unselected categories display colored box only (text hidden for privacy) - Unassigned entries can be shown/hidden using the unassigned filter marker - Visibility state is shared with Tasks page via settings persistence ## [0.9.773] - 2025-12-25 ### Changed - Checklist Correction UX: Corrections now use a delayed-save pattern with cancel option - When editing a checklist item triggers a correction, a snackbar appears with a 5-second countdown - Users can tap "CANCEL" to discard the correction before it's saved - If the countdown completes without cancellation, the correction is saved to the database - Prevents accidental one-time fixes from being permanently saved as AI learning examples ## [0.9.772] - 2025-12-24 ### Added - AI Linked Task Context: AI prompts now include context from related tasks - `{{linked_tasks}}` placeholder provides parent/child task relationships - Includes task metadata (status, priority, time spent, labels) for each linked task - Includes latest AI summary for linked tasks (with GitHub/external link awareness) - Batched database queries avoid N+1 performance issues - Labels resolved via O(1) cache lookup ### Changed - AI Repository: Decoupled prompt semantics from data construction - Note about web search for external links moved from repository to prompt builder - Repository now returns pure data; prompt builder adds contextual instructions ## [0.9.771] - 2025-12-23 ### Added - Task Summary Link Extraction: AI-generated task summaries now include a "Links" section - AI is instructed to scan log entries and extract unique URLs (http://, https://, etc.) - Generates succinct, descriptive titles for each link (e.g., "Linear: APP-123", "Lotti PR #456") - Formats as clickable Markdown links - Links section omitted when no URLs are found (prompt-driven, best-effort) ## [0.9.770] - 2025-12-23 ### Changed - Upgraded to Flutter 3.38.5 and updated dependencies - Sync: Refactored `OutboxService.enqueueMessage()` for improved maintainability - Converted from if-else chains to Dart 3 pattern matching with exhaustive switch - Extracted `_enqueueSimple` shared helper to reduce code duplication across 7 simple message handlers - Extracted per-type handler methods (`_enqueueJournalEntity`, `_enqueueEntryLink`, etc.) - Added preparation methods (`_prepareJournalEntity`, `_prepareEntryLink`) for message pre-processing - Improved test coverage: 60 → 77 tests with edge case coverage for null values and error handling ### Fixed - Tests: Fixed resource leak in OutboxService tests where auxiliary service instances were not disposed ## [0.9.768] - 2025-12-23 ### Added - Gemini Thinking Support: All Gemini 2.5+ models (including Flash) now support thinking/reasoning - Thoughts are displayed in the AI Response Summary Modal's "Thoughts" tab - Removed incorrect suppression of thinking for Flash models - Usage Statistics: Track and display AI inference metrics - Input tokens, output tokens, and thoughts tokens from providers - Processing duration measurement in milliseconds - Statistics displayed in AI Response Summary Modal - Conversation Manager: Added thought signature storage for future multi-turn function calling support ### Changed - Standardized thinking block format from `` to `` tags for consistency across providers ## [0.9.765] - 2025-12-21 ### Fixed - Sync: Remove live-scan look-behind tail logic and the `enable_matrix_lookbehind_tail` flag so live scans always process strictly-after slices only. ## [0.9.764] - 2025-12-21 ### Fixed - Checklist Deletion: Fix persistent empty card after checklist deletion - When deleting a checklist, the checklist ID is now removed from the parent task's `checklistIds` list - Added defensive UI filtering to handle existing stale checklist references - Empty card containers no longer appear after checklist deletion ## [0.9.762] - 2025-12-19 ### Fixed - Sync Catch-Up: Prevent concurrent catch-up execution - Added `_forceRescanInFlight` guard to serialize concurrent `forceRescan()` calls (e.g., connectivity + startup handlers firing simultaneously) - Added `_catchUpInFlight` check in `forceRescan()` to skip catch-up when another is already running from `_runGuardedCatchUp()` or `_scheduleInitialCatchUpRetry()` - Fixes issue where catch-up from external triggers could run concurrently with internal retry-driven catch-ups, causing `processOrdered` timeout failures - `bypassCatchUpInFlightCheck` parameter allows internal callers like `_startCatchupNow()` to bypass the check when they've intentionally set the flag ## [0.9.761] - 2025-12-18 ### Fixed - Sync Catch-Up: Wait for SDK sync completion before running catch-up - All catch-up paths (startup, app resume, wake, reconnect) now wait up to 30 seconds for the Matrix SDK to complete its `/sync` with the server - Fixes issue where catch-up after extended offline periods would only retrieve a few entries instead of the full backlog - If timeout occurs on slow networks, a follow-up catch-up automatically triggers when sync eventually completes - Tuning: `SyncTuning.catchupSyncWaitTimeout` (default 30s) ## [0.9.758] - 2025-12-18 ### Fixed - Sync Gap Detection: Prevent false positives during reconnection - Covered vector clocks now processed BEFORE gap detection, preventing intermediate counters from being incorrectly marked as missing - Live scan signals are deferred while catch-up is processing older events, ensuring in-order ingest and preventing newer events from triggering false gaps before older events are recorded ## [0.9.755] - 2025-12-16 ### Added - Nested AI Responses: Display generated prompts directly under audio entries - AI responses linked from audio entries now appear in an expandable section - Collapsible UI with smooth animation to show/hide nested responses - Swipe-to-delete with confirmation dialog for removing AI responses - Error feedback when deletion fails ## [0.9.754] - 2025-12-15 ### Added - Sync Backfill Improvements: EntryLink support and ghost entry resolution - EntryLink messages now tracked in sequence log alongside journal entities - New `SyncSequencePayloadType` enum distinguishes `journalEntity` vs `entryLink` - Ghost missing entry resolution: when different payload types share the same sequence counter, receiving one resolves the other (e.g., EntryLink at `(alice:5)` resolves a missing JournalEntity counter) - `populateFromEntryLinks` method populates sequence log from existing entry links - Extracted `SequenceLogPopulateProgress` widget for testable progress UI - Comprehensive test coverage for backfill response handler, sync event processor, sync database, and outbox service ## [0.9.752] - 2025-12-06 ### Added - Sync Backfill: Request missing journal entries from connected devices during sync - Automatically detects gaps in local journal entries based on sync status - Requests missing entries from peers to ensure complete synchronization - Integration tests verify backfill functionality ### Fixed - Sync: Treat missing attachment fetches as retryable failures so markers do not advance past incomplete entries, serialize `forceRescan` with a completer to avoid overlapping runs, and include Matrix stream consumer instance IDs in sync logs to track concurrent pipelines. - Sync: Avoid redundant descriptor catch-up retries by running catch-up only on pending changes, triggering retryNow only when new descriptors are discovered, and skipping stale descriptor errors when the local entry already supersedes the incoming vector clock. - Sync: Skip no-op entry link updates so vector clocks only advance when link state changes. - Sync: Preserve sequence log `createdAt` when resolving covered counters to avoid upsert validation errors. - Sync: Wait for in-flight ordered processing to finish instead of timing out catch-up batches during long attachment downloads. - Sync: Queue attachment downloads asynchronously (bounded concurrency) so ordered processing does not block on large media. - Sync: Skip duplicate journal entity messages with the same vector clock to reduce redundant older-or-equal applies. - Sync: Send descriptor JSON from a single snapshot so the message vector clock matches uploaded bytes, avoiding stale vector-clock retries. - Sync: Mark superseded own counters as unresolvable during backfill so requested entries can clear. - Sync: Always include the current vector clock in `coveredVectorClocks` for journal entries and entry links while ignoring the current clock for pre-marking to preserve gap detection. - Sync: Fill missing `originatingHostId` on outgoing journal entities and entry links so sequence logs always update. - Outbox: Respect retry backoff across triggers to prevent rapid retry storms under flaky networks. - Tests: Stabilize Matrix stream consumer signal coverage around live scan timing and stubs. ### Changed - Sync: Split the Matrix stream consumer into catch-up, live-scan, processing, and signal components, and split the pipeline tests into smaller files for maintainability. ### Added - Automatic Image Analysis: Images added to tasks are now analyzed automatically - When dropping, pasting, or importing images to a task with image analysis enabled, analysis runs in background - Uses category's `automaticPrompts[imageAnalysis]` configuration - Fire-and-forget pattern ensures image import is never blocked - Platform-aware: automatically selects available prompts for current platform - New `AutomaticImageAnalysisTrigger` helper class with `onCreated` callback in `JournalRepository.createImageEntry()` - Tests: 10 unit tests with 100% coverage - See: `docs/implementation_plans/2025-12-01_automatic_image_analysis_and_task_summary_triggers.md` - Smart Task Summary Triggers: Simplified and unified task summary creation - First summary created immediately when meaningful content is added (if auto-summary enabled) - Subsequent updates use existing 5-minute countdown mechanism - Triggers on: image analysis completion, audio transcription completion, manual text save (non-empty) - Avoids "lame" first summaries by requiring meaningful content before triggering - New `SmartTaskSummaryTrigger` helper class with dual-path logic - Integrated into `UnifiedAiInferenceRepository._handlePostProcessing()` and `EntryController.save()` - Tests: 9 unit tests with 100% coverage + 5 entry controller tests - See: `docs/implementation_plans/2025-12-01_automatic_image_analysis_and_task_summary_triggers.md` - Checklist Correction Examples: Learn from manual corrections to improve AI accuracy - When users manually correct a checklist item title, the before/after pair is captured - Examples are stored per-category and injected into AI prompts via `{{correction_examples}}` - Supports both checklist update prompts and audio transcription prompts - New `CategoryCorrectionExamples` widget in category settings with swipe-to-delete - Smart filtering: ignores no-change, trivial case-only, and duplicate corrections - Warning banner when approaching token budget (400+ examples, max 500 injected) - Auto-dismissing snackbar confirms successful correction capture - Fire-and-forget capture via `unawaited()` for zero UI latency - Syncs across devices via existing category sync infrastructure - See: `docs/implementation_plans/2025-11-30_checklist_item_correction_examples.md` - Speech Dictionary per Category: Improve transcription accuracy with domain-specific terms - Categories can now store a speech dictionary of correct spellings for names, places, and technical terms - Dictionary terms are injected into AI transcription prompts via `{{speech_dictionary}}` placeholder - New `CategorySpeechDictionary` widget in category settings (semicolon-separated text field) - Context menu in QuillEditor: select corrected text and "Add to Speech Dictionary" - Works for tasks, linked audio entries, and linked image entries - `SpeechDictionaryService` handles term addition with validation (max 50 chars, trimming) - Terms sync across devices via existing category sync infrastructure - Tests: 22 service tests, 16 widget tests, 14 prompt builder tests - See: `docs/implementation_plans/2025-11-30_speech_dictionary_per_category.md` - AI Task Summary: Scheduled refresh with countdown UX - Changed from 500ms debounce to 5-minute scheduled delay - UI shows countdown "Summary in 4:32" when refresh is scheduled - Cancel button (✕) to stop scheduled refresh - Trigger Now button (▶) to bypass countdown and generate immediately - Additional checklist changes batch into existing countdown (no timer reset) - Reduces API costs while actively working on a task - Implementation: `DirectTaskSummaryRefreshController` returns `ScheduledRefreshState` - Efficient countdown updates via `StreamBuilder` in `_HeaderText` widget - Tests: 15 unit tests + 7 widget tests for scheduled refresh behavior - AI – Checklist Updates improvements - Current Entry hint: recording modal and entry-level AI popup now pass the focused entry ID so prompts prioritize the user-edited transcript/text before falling back to the full task log. - Deleted items context: prompt builder injects every soft-deleted checklist title (with deletion timestamps) so the LLM can avoid recreating them. - `update_checklist_items` tool: unified update path for existing checklist items, supporting: - Mark items as checked/unchecked (e.g., "I did X" → item marked complete) - Fix transcription errors in titles (e.g., "mac OS" → "macOS", "i Phone" → "iPhone") - Combined updates: status and title correction in a single call - Semantic matching: AI matches user references to items by meaning, not exact text - Popup parity: running Checklist Updates from a linked audio/image entry now threads that entry as `linkedEntityId`; task-level runs keep the parameter unset for whole-task analysis. - Checklists: Ergonomics improvements - TitleTextField: Cmd/Ctrl+S saves; add-item field retains focus after save for rapid entry - Filter: default “Open only” with header toggle and completed count (N/M done) - Accessibility: keyboard toggle (Cmd+Shift+H / Ctrl+H) + screen reader announcements - Persistence: per‑checklist filter mode is remembered - Empty state: “All items completed!” message when open‑only has no items - Checklists: Completion feedback animation - Checking off an item now plays a subtle green-tinted border/glow “fanfare”. - In “Open only” view, completed items fade out and collapse over a short duration instead of disappearing instantly. - In “All” view, items keep the highlight but remain visible so users can still see what was just completed. - Journal/Tasks: Active timer highlight — linked entries that match the running timer now render with a persistent red glow for quick visual identification. - Journal/Tasks: Temporary scroll highlight — after auto-scrolling to a linked entry, the target briefly glows to confirm focus. - AI label assignment: Category-scoped label suggestions (Phase 1 guardrails) - AI only suggests labels applicable to the task's category (global ∪ scoped labels) - Prompt includes category-filtered labels; ingestion enforces category scope - Telemetry tracks `out_of_scope` skip reason with full structured payload - TaskContext parameter reduces DB lookups when category is already available - Tasks: Priority (P0–P3) end-to-end support - New TaskPriority field with compact P0–P3 chips and colors - Header picker + detail chip; selection modal uses the chip visuals - Task list filtering by priorities and ordering by priority rank, then date - Persistence: priority filter stored per-tab with other task filters - Migration v29: adds `task_priority` (TEXT) and `task_priority_rank` (INTEGER) with backfill to P2/2 for legacy tasks and updated composite index - i18n: English keys for labels, picker title and descriptions - Tests: database ordering/filtering, UI wrappers, filter UX - Checklists (Open-only filter): Newly completed items no longer disappear instantly or flicker when toggling between “All” and “Open only”; rows now remain visible briefly with a completion animation before smoothly fading out. ### Fixed - **Sync: Entry link atomicity** — Calendar entries no longer appear grey due to missing category information - Entry links now embedded within journal entity sync messages, ensuring atomic processing - Links only processed after successful entry persistence, preventing orphaned references - Backward compatible: standalone entry link messages still supported for link updates - Reduces sync message count from N+1 to 1 per entry with links - Comprehensive logging: `enqueueMessage.attachedLinks` and `apply.entryLink.embedded` events - Implementation: `lib/features/sync/model/sync_message.dart`, `lib/features/sync/outbox/outbox_service.dart`, `lib/features/sync/matrix/sync_event_processor.dart` - Tests: 3 new tests in `test/features/sync/outbox/outbox_service_test.dart` - See: `docs/implementation_plans/2025-11-16_entry_link_sync_atomicity.md` - Tasks page: Active label filters are now visible below the search header (no longer clipped in the app bar). - AI label assignment: Prevented out-of-category labels from being assigned by AI - Task label selector: Now shows currently assigned out-of-scope labels to allow unassigning - List is strictly A–Z (case-insensitive); selection does not change ordering - Out-of-category assigned labels are included with an "Out of category" note when applicable - Solves issue where out-of-scope labels were hidden and couldn't be removed - Label creation UX: "Create label" option now available with substring matches - Previously hidden when typing "CI" while "dependencies" existed (substring match) - Now shows "Create 'CI' label" button below filtered results when no exact match exists - Only hides create option when exact match (case-insensitive) is found - Improves discoverability for short label names that appear as substrings - Sync maintenance now supports Labels in the manual definition sync flow: - Sync page → Sync Entities modal includes a "Labels" checkbox alongside Tags, Measurables, Categories, Dashboards, Habits, and AI Settings. - Each step reports per-step progress and totals; selections can be run together. - tests(labels/ai): - Max-volume labels selection (500+) with performance and ordering checks - Prompt injection protection for label names - Conversation retry and interruption coverage for `assign_task_labels` - TaskLabelsWrapper toast + undo + rapid assignment behavior - LabelValidator validity/concurrency tests - Summary note test when labels exceed the prompt cap - AI label assignment with function-calling (see PR #2365): - Prompt label injection (capped to 100, usage + alphabetical) with optional private filtering. - `assign_task_labels` tool (add-only) with max 5 labels per call. - Shared rate limiter (5-minute window) and shadow mode. - Flatpak packaging: install desktop and AppStream metadata from the upstream repo and use a `type: script` wrapper instead of inline `cat` heredocs in the manifest. - User feedback: SnackBar in Task Details listing assigned labels with Undo to remove them. - Event stream for UI notifications and comprehensive unit/feature tests. - Settings header widget tests now cover multiple text scale factors (1.0, 1.2, 1.5, 2.0) and common screen widths to guard layout across devices. - Label chips now surface tooltips/long-press descriptions, Task label sheets show inline "Create label" CTAs, and the journal header includes a quick label filter row for active selections. - Labels settings list displays usage counts sourced from a new `watchLabelUsageCounts` stream plus widget coverage for the editor sheet, list page, assignment sheet/wrapper, and filter chips. - Added integration (`label_workflow_test.dart`), accessibility, repository edge-case, and performance tests (1k+ tasks + reconciliation benchmarks) for the labels system. - AI prompt selection modal now visually highlights default automatic prompts with gold accent (border and icon background). - Platform-aware AI prompt filtering automatically hides local-only models (Whisper, Ollama, Gemini 3N) on mobile platforms. - Fallback logic ensures default automatic prompts gracefully switch to available alternatives when local-only models are filtered on mobile. - Comprehensive test coverage added for platform filtering, isDefault prompt highlighting, and ModalCard border/animation behavior. - Outbox: Pause processing while logged out and surface a one-time red toast ("Sync is not logged in") when sync is enabled but not authenticated. Prevents wasted retries and clarifies state to the user. - Labels: Applicable categories (scoped labels) - New optional `applicableCategoryIds` on `LabelDefinition` (backward compatible JSON only) - Reactive provider for category-scoped availability using cache + streams - EntitiesCacheService builds `global` and per-category buckets with pruning on category changes - Label editor adds an "Applicable categories" section (chips + add/remove) - Task label picker filters to union of global + current category - Repository validates category IDs, de-dupes and sorts by category name for stable diffs - Tests: service unit tests and widget tests updated; i18n keys added with missing translations noted ### Changed - AI: Task-related prompts default to non-streaming, with a new `enable_ai_streaming` flag controlling streaming for those actions (chat remains streaming). - Flatpak: AppStream metadata version/date now come directly from the checked-in `com.matthiasn.lotti.metainfo.xml`; the manifest tool no longer rewrites them during Flathub prep. - Flatpak: Fail fast when flatpak-flutter or cargo source generation fails (no fallback paths), keep tool PATHs via append-path (including `/usr/bin:/bin`), and keep downloaded Cargo.lock files only in the working directory (not in submission artifacts) for reproducible builds. - Journal/Tasks: Refactored scroll-to-entry and highlight logic into `HighlightScrollMixin` with configurable durations and a small retry backoff; `LinkedEntriesWithTimer` scopes rebuilds to the linked entries section only. Focus intent is cleared early (next frame) for responsiveness and also on success/terminal failure for robustness. - Tasks UI: Checklist item vertical spacing reduced (outer padding 2px; content padding vertical 2px) for a more compact list. - UI/Tasks: Redesigned the active label filters header below the search bar - Compact card-style container (rounded, `surfaceContainerHighest`) with smooth `AnimatedSize` - Header shows a subtle filter icon and “Active label filters (n)” - Clear action uses compact `TextButton.icon` (smaller text) and chips use compact density - Renders only when filters are active; no empty container state - Sync (Outbox): enforce a strict idle window (3s) with no forced 2s deadline; pause drains mid-burst when activity resumes and reschedule via the standard retry delay. - Sync (Matrix): Client stream is now signal-driven and always triggers a catch-up via `forceRescan(includeCatchUp=true)` with an in-flight guard to prevent overlaps. Timeline callbacks continue to schedule debounced live scans and fall back to `forceRescan()` on scheduling errors. - Sync (Matrix): Documentation refreshed to reflect signal-driven ingestion and backlog completion behavior (`lib/features/sync/README.md`, `docs/sync/sync_summary.md`). - Sync (Matrix): Coalescing and throttling refinements - Catch-up coalesces signals with a 1s minimum gap and runs exactly one trailing pass after bursts; logs `catchup.start` once and `catchup.done events=…` once per burst. - Live-scan never overlaps; signals during a scan defer as a single trailing pass. Base debounce ~120ms is extended to honor a 1s min gap; logs `signal.liveScan.coalesce` and `trailing.liveScan.scheduled`. - Double-scan for attachments now awaits the immediate second pass; a delayed pass runs at +200ms. - Historical windows reduced: catch-up `preContext=80`, `maxLookback=1000`; live-scan steady tail=30; audit tails 50/80/100. - Log volume reduced under `collectMetrics`; condensed `marker.local` to id+ts. - feat(ai/labels): Append a summary note after the labels JSON in prompts when the number of available labels exceeds the cap, e.g. `(Note: showing 100 of 150 labels)`. - Matrix Sync Stats page now uses the modern SettingsPageHeader with collapsing sliver layout and subtitle, aligning with the new settings header UX. - Extracted header spacing and breakpoints into `lib/widgets/app_bar/settings_header_dimensions.dart` for maintainability and consistency. - Settings header filter card now hugs the count summary by trimming the extra padding underneath it. - Tasks and journal tabs now persist category filter selections independently, restoring their own state after app restart. - Category filter storage migrated from shared `TASK_FILTERS` key to per-tab keys (`TASKS_CATEGORY_FILTERS` and `JOURNAL_CATEGORY_FILTERS`). - Task status filters (Open, In Progress, Done, etc.) remain scoped exclusively to the tasks tab. - AI prompt availability now respects platform capabilities, preventing confusion from unusable model options on mobile. ### Fixed - Labels: description clearing when deleting the last character - Controller sanitizes input (trim + remove NBSP/ZWSP/BOM) and stores `null` in state when empty. - On update, the controller signals a clear by sending `''` (empty string) instead of `null`. - Repository update semantics: `null` = unchanged, empty string → clear (persist as `null`), non‑empty → trimmed. - Prevents the last stray character from reappearing after Save when the field was cleared. - Tests added for repository clearing and Label Details UI (keyboard shortcuts, cancel, error rendering, privacy toggle, category chip removal, controller reseed behavior). Codecov patch coverage improved for `label_details_page.dart` and `labels_list_page.dart`. - AI: Hardened checklist item parsing to avoid accidental splitting when items contain commas or grouped text: - `add_multiple_checklist_items` now accepts a JSON array of strings (preferred) or a robustly parsed string. - Escaped commas (`\,`), quoted items ("..." / '...'), and commas inside parentheses/brackets/braces no longer split into separate items. - Updated prompt guidance to prefer arrays; fallback string format documented with escaping. - Added unit tests for parser and batch handler; analyzer warnings resolved. - Sync Outbox: The red "Sync is not logged in" toast now triggers only when an outbox send is attempted while logged out, not on app startup. Prevents noisy startup toasts when users haven’t logged in yet. - Sync (Matrix): Catch-up now continues escalating snapshot size until it’s not full (or lookback cap reached), ensuring the entire backlog after the read marker is retrieved. This eliminates missing EntryLinks after offline windows and prevents gray boxes on return to online. - Sync (Matrix): Fixed overlapping live-scans by guarding `_scanInFlight` with a depth counter and awaiting the immediate pass in attachment double-scans. - Sync Outbox: Eliminated "stuck after reconnect" cases by adding a watchdog (10s), DB nudge on outbox count changes (50ms), send timeout (20s) with `timedOut=true` logging, and pass-cap continuation; ClientRunner now catches callback errors and LoggingService DB failures are best-effort. - Stabilized labels/task widget tests by awaiting `getIt.reset()`, providing scoped service mocks, and giving sheet/editor hosts real `MediaQuery` sizes so chips, toggles, and CTAs are tappable during automation. - Waveform: show visualization for recordings longer than 3 minutes by removing the duration gate and introducing dynamic zoom scaling for long clips; updated tests and docs accordingly. - Tasks: aligned task header date/progress typography and ensured mobile task list cards hide HH:MM progress text. ### Changed - AI (checklist): Unified to array‑only batch creation. The single‑item tool is no longer used in conversations; tests and prompts now require `{ "items": [{"title": "...", "isChecked"?: bool}] }`. - AI (tests): Stabilized checklist conversation tests by stubbing `ConversationRepository.sendMessage` to invoke `LottiChecklistStrategy.processToolCalls(...)` with predefined tool calls, avoiding brittle streaming mocks. - AI (tests): Removed legacy skipped tests in `test/features/ai/functions/lotti_conversation_processor_test.dart` in favor of repo‑wired cases in `test/features/ai/functions/lotti_conversation_processor_via_repo_test.dart`. ### Tests - Added deterministic scenarios: non‑GPT‑OSS batch, single via batch, 3‑item batch, GPT‑OSS batch, and language detection → create. All pass reliably. - Hardened `lotti_batch_checklist_handler_test.dart`: assert non‑null item creation when appending to existing checklists; verify callback and payload; analyzer clean. ### Documentation - Updated `lib/features/ai/README.md` with deterministic conversation testing guidance and file references. - Expanded inline docs in `lib/features/ai/conversation/conversation_repository.dart` and `lib/features/ai/functions/lotti_conversation_processor.dart` to document streaming expectations and deterministic testing. ## [0.9.704] - 2025-10-24 ### Added - Comprehensive Fts5Db coverage with insert and search stream tests to guarantee new database behaviour. - Added extensive database test coverage for purge flows, conversion utilities, sync outbox edge cases, logging safeguards, and settings persistence. ### Changed - Retired the legacy Matrix sync V1 pipeline in favour of the stream-first implementation. ## [0.9.703] - 2025-10-24 ### Added - Timer indicator now auto-scrolls to the running entry when tapped from task details page. - Task focus controller for managing scroll-to-entry intent across navigation events. ### Changed - Expanded speech waveform coverage with scrubber interaction tests, painter edge cases, cache I/O failure guards, and path sanitisation scenarios. ### Fixed - Stabilised waveform cache pruning tests by clearing mock interactions between scenarios. ## [0.9.702] - 2025-10-23 ### Changed - Matrix sync now tracks locally emitted Matrix event IDs in a shared sent-event registry, short-circuiting echo events during ingestion and recording suppression metrics to cut redundant network and database work. ## [0.9.701] - 2025-10-23 ### Added - Support Markdown divider insertion from the journal editor toolbar. ## [0.9.700] - 2025-10-23 ### Changed - Split the smart journal loader into dedicated `DescriptorDownloader` and `VectorClockValidator` components so caching, purging, and vector-clock logic are independently testable. ### Fixed - Matrix V2 catch-up now advances markers after offline sessions by correcting journal update result semantics, eliminating false "missing base" retries, and trimming duplicate backlog processing. - Matrix outbox refreshes journal JSON before enqueueing and Matrix sender re-syncs vector clocks with descriptor payloads, preventing stale checklist descriptors from being uploaded. - Added a circuit breaker so stale descriptor downloads bail out after repeated refresh attempts instead of looping forever. - Bumped patch version to 0.9.700+3390 to ship the sync catch-up contract fix without breaking APIs. ## [0.9.699] - 2025-10-22 ### Changed - Sync filter chips now hide non-informational zero badges, emphasise pending/error/unresolved totals with tinted badges, trim their height slightly, and keep empty-state cards constrained on the conflicts page. ## [0.9.698] - 2025-10-22 ### Changed - Audio player progress bar now matches the play/pause ring by using `ColorScheme.primary` for both the scrub fill and contrast-adjusted playhead, softens the playback card shadow, and removes the glow to avoid black playheads in high-contrast themes. Light mode swaps the gradient background for a flat surface tint to keep the card crisp. ## [0.9.697] - 2025-10-22 ### Changed - Speech audio playback card now ships with glassmorphism styling, streamlined controls, custom progress bar, and dedicated widget coverage for play/pause and speed interactions. ## [0.9.696] - 2025-10-21 ### Changed - Moved Matrix sync maintenance actions (delete sync database, re-sync definitions/messages) to the new Matrix Sync Maintenance page under Sync Settings. - Sync Outbox and Sync Conflicts list pages now use modern cards with segmented filters, inline counts, and polished empty states via the shared sync list scaffold. - Added dedicated widget tests for `ConflictListItem` and `SyncListScaffold` covering filters, semantics, and interaction paths. ### Fixed - Matrix Stats `Last updated` label now stays stable when metrics payloads are unchanged, eliminating refresh jitter. - Guarded sync list filters against invalid persisted enum indexes on conflicts/outbox pages to prevent `RangeError`. ## [0.9.695] - 2025-10-21 ### Changed - Matrix Stats now keeps per-type "Sent" counts stable while the page stays flicker-free on mobile and desktop. - Read markers no longer spam `M_UNKNOWN` errors—local-only IDs are skipped and expected misses are logged once. ## [0.9.694] - 2025-10-21 ### Changed - Sync UX overhaul: - Promote Sync to a top-level Settings entry (hidden when Matrix sync flag is off) - Move Outbox Monitor under `/settings/sync/outbox` and remove redundant toggle - Surface Matrix Stats as a full page under `/settings/sync/stats` with improved loading state - Clean up Advanced: remove Matrix/Outbox/Conflicts tiles; keep Logs, Health Import (mobile), Maintenance, About - Route matching for Sync pages now uses exact path matching for robustness - Localize new user-facing strings (tile subtitles and error text) ## [0.9.693] - 2025-10-21 ### Changed - Hid pre-release features when their corresponding tabs are disabled, tightening entry-type gating and extending localisation coverage. ## [0.9.692] - 2025-10-20 ### Changed: - Enforce tabular monospace style for tasks timers (monoTabularStyle) to eliminate width jitter. ## [0.9.691] - 2025-10-20 ### Changed: - Audio recorder: improve prompt checkbox visibility for better contrast and readability ## [0.9.690] - 2025-10-20 ### Added: - Export checklist as Markdown from the checklist view (#2331) - Share checklist via the iOS/macOS share sheet (#2331) ## [0.9.689] - 2025-10-19 ### Changed: - Upgraded dependencies across the project (#2330) ## [0.9.688] - 2025-10-19 ### Changed: - Improve initial synchronization reliability and speed at app start (#2329) ## [0.9.687] - 2025-10-17 ### Changed: - Harden v2 sync with stronger error handling and state management (#2327) ## [0.9.686] - 2025-10-14 ### Fixed: - Clamp bottom navigation index to a valid range to avoid out-of-bounds selection (#2323) ## [0.9.685] - 2025-10-14 ### Added: - Drag and drop audio to quickly add or import audio content (#2322) ## [0.9.684] - 2025-10-13 ### Changed: - Refined sync stats UI for clearer labels and improved readability (#2321) ## [0.9.682] - 2025-10-12 ### Added: - Copy as plain text from supported views (#2319) - Copy as Markdown with formatting preserved (#2319) ## [0.9.680] - 2025-10-12 ### Changed: - Improve sync reliability with better retry and state handling (#2317) ## [0.9.679] - 2025-10-12 ### Changed: - Upgraded Matrix SDK to 3.0.0 and adapted usage where necessary (#2316) ## [0.9.678] - 2025-10-12 ### Added: - Emoji input and rendering support on Linux (#2318) ## [0.9.677] - 2025-10-10 ### Fixed: - Issue where sync could lag one event behind (#2313) ## [0.9.676] - 2025-10-09 ### Changed: - Internal sync code cleanup and refactoring to reduce complexity (#2308) ## [0.9.675] - 2025-10-07 ### Changed: - Extract session, room, and timeline managers into dedicated components (#2304) ### Fixed: - Auto-join bug in Matrix integration (M6) (#2304) ## [0.9.674] - 2025-10-06 ### Added: - Exponential backoff for sync retry strategy to improve stability on flaky networks (#2299) ## [0.9.673] - 2025-10-06 ### Changed: - Sync maintenance refactor and tidy-up of helpers and lifecycles (#2300) ## [0.9.672] - 2025-10-06 ### Changed: - Improve sync login resilience with better recovery from transient errors (#2298) ## [0.9.671] - 2025-10-05 ### Fixed: - Clipped text in editor (#2293) ## [0.9.670] - 2025-10-04 ### Changed: - Update task summaries when changing the app language to keep content in sync (#2291) ## [0.9.669] - 2025-09-27 ### Changed: - Upgraded dependencies ## [0.9.668] - 2025-09-23 ### Added: - Gemma 3n local AI provider support for audio transcription and text generation - Streaming support for Gemma 3n text generation with OpenAI-compatible API - Internationalization for Gemma 3n provider strings ### Changed: - Refactored provider configuration to centralize API key requirement logic - Improved JSON response validation in Gemma 3n repository with try-catch error handling - Hide API key field for local providers (Ollama, Whisper, Gemma 3n) ## [0.9.667] - 2025-09-22 ### Added: - Gemma 3N audio transcription service with Docker and local Python support - Self-contained Flatpak build system with automated submission scripts - Improved screenshot portal service for Flatpak environments ### Changed: - Updated app identifier from com.matthiasnehlsen.lotti to com.matthiasn.lotti across all platforms - Enhanced Gemma service with improved error handling and streaming capabilities - Upgraded Flutter to version 3.35.4 - Upgraded dependencies ### Fixed: - Gemma service transcription improvements with better audio processing - Flatpak build process streamlining and reliability enhancements ## [0.9.665] - 2025-09-17 ### Changed: - Update app identifier to use with different Apple developer account ## [0.9.664] - 2025-09-17 ### Changed: - Upgraded Flutter to version 3.35.4 - Upgraded dependencies (minor versions) ## [0.9.663] - 2025-09-14 ### Changed: - const factories - Upgraded Matrix SDK to version 2.0.1 - Improved `deleteDevice` method with proper validation and error handling ### Fixed: - Replaced deprecated `loginState` with `client.isLogged()` method - Added proper password authentication for device deletion - Added user feedback for device deletion operations with success/error messages ### Added: - Validation for device ownership before deletion attempts - Clear error messages when device deletion fails - TODO for future SSO/token authentication support in device deletion ## [0.8.393] - 2023-06-29 ### Changed: - Consistent microphone icon across app - Upgraded dependencies ### Fixed: - Sort categories - Bottom sheet heights ## [0.8.392] - 2023-06-24 ### Changed: - Hide measurement suggestions when input dirty - Upgraded dependencies ## [0.8.391] - 2023-06-24 ### Changed: - Keyboard dismiss behavior ## [0.8.390] - 2023-06-23 ### Added: - Log sync email subject for debugging ## [0.8.389] - 2023-06-23 ### Changed: - Smoother scrolling infinite journal page ## [0.8.388] - 2023-06-23 ### Changed: - Upgraded dependencies ### Fixed: - Text color on tag and task status chips ## [0.8.387] - 2023-06-22 ### Changed: - Unfocus keyboard on entry/task save ## [0.8.386] - 2023-06-22 ### Fixed: - Habits search field style ## [0.8.385] - 2023-06-22 ### Changed - Delete icon style - Start new tag dialog with search field content - Whitespace on entity detail pages ## [0.8.384] - 2023-06-21 ### Changed - Audio recorder and player icons and style ## [0.8.383] - 2023-06-21 ### Changed: - Upgraded dependencies ## [0.8.382] - 2023-06-19 ### Added: - Shimmer animation on habit success button ### Changed: - Upgraded dependencies - Choice chip styles - Habit skip button saturation - Smoother transitions on navigate ## [0.8.381] - 2023-06-17 ### Changed: - Typography in DateTime modal - Selected color in editor toolbar ### Fixed: - Survey next button text color ## [0.8.380] - 2023-06-16 ### Fixed: - Border in search widget ## [0.8.379] - 2023-06-16 ### Changed: - Update tasks page when task created ## [0.8.378] - 2023-06-16 ### Fixed: - Update whisper model size after completed download ## [0.8.377] - 2023-06-15 ### Changed: - Upgraded icons - Consistent text style in input fields - Prominent name fields in entity definitions - New Flutter version - Upgraded dependencies ### Fixed: - Text weight and whitespace in entry card footer - Form field text color in light mode - Flaky test - Habit skip and completion colors - Clipped habit bottom sheet corners ## [0.8.376] - 2023-06-12 ### Changed: - Text weight and whitespace in entry card footer - Splash screen ### Fixed: - Filter choice chip text color - Short display of black screen at startup - Missing routes - Not found error on tasks page ## [0.8.375] - 2023-06-10 ### Fixed: - App data folders on Linux ### Changed: - Upgraded dependencies ## [0.8.374] - 2023-06-09 ### Changed: - Refactoring: remove ThemesService - Settings page for light and dark color themes ## [0.8.373] - 2023-06-08 ### Fixed: - BP chart header ## [0.8.372] - 2023-06-08 ### Changed: - Improved dashboards page header - New Flutter version ## [0.8.371] - 2023-06-08 ### Changed: - Colors in measurables dialog - Borders in bottom sheets and dialogs - Upgraded dependencies ## [0.8.370] - 2023-06-07 ### Changed: - Upgraded dependencies - Non-persistent title sliver app bar on dashboards page - Spacing in habits search header - Habit completion card typography & whitespace ### Fixed: - Keyboard brightness when following system dark/light modes ## [0.8.369] - 2023-06-06 ### Changed: - Improve habit completion chart colors ## [0.8.368] - 2023-06-06 ### Added: - Config flag for following system brightness ## [0.8.367] - 2023-06-05 ### Changed: - Color schemes from flex_color_scheme - Styling - Style fixes & refactoring - Styles refactoring - Text editor border - Editor toolbar elevation - Habit completion bottom sheet background styling ### Fixed: - Task duration estimate bottom sheet - Font colors on survey bottom sheet - Color issues - Charts layout - Colors in dashboard chart selector - Colors in entry delete bottom sheet - Colors in measurables definition page - Tag search text color ## [0.8.366] - 2023-06-03 ### Changed: - Adapt filter chips to match Material 3 design - Upgraded dependencies & Flutter upgrade - Use segmented button with icons for filtering journal ## [0.8.365] - 2023-06-02 ### Fixed: - Display of Whisper model sizes ## [0.8.364] - 2023-06-01 ### Changed: - Use habit completion card as habit chart in dashboard ## [0.8.363] - 2023-06-01 ### Changed: - Upgraded dependencies - Time span segmented controls style to match habit statuses ## [0.8.362] - 2023-05-31 ### Fixed: - Category opacity ### Changed: - License: GPL - Upgraded dependencies ## [0.8.361] - 2023-05-31 ### Changed: - Habit status segmented control style - Upgraded dependencies ## [0.8.360] - 2023-05-30 ### Changed: - Upgraded dependencies - Health import in transaction ### Fixed: - Locked db when importing from health ## [0.8.359] - 2023-05-26 ### Fixed: - Default language English instead of German (first in alphabetical order) ### Changed: - Flutter upgraded to 3.10.2 & upgraded dependencies ## [0.8.358] - 2023-05-23 ### Fixed: - Crashes when attempting simultaneous transcriptions ## [0.8.357] - 2023-05-22 ### Changed: - Upgraded dependencies - Larger map with interaction ### Added: - Display size of downloaded whisper models ## [0.8.356] - 2023-05-21 ### Changed: - Improve header style on settings page using Slivers - Improve header style on advanced settings page using Slivers - Health import styling - Hide health import page on all platforms but iOS - Upgraded dependencies - Add back button on advanced settings page - Sliver app bar in entity definitions - Sliver app bar in speech settings - Sliver app bar in config flags page - Sliver app bar on maintenance page - Sliver app bar on about page - Sliver app bar on health import page - Sliver app bar on dashboard page - Sliver app bar on logging page ### Added: - Tests for category settings page - Tests for habit filters - Tests for dashboards page ## [0.8.355] - 2023-05-19 ### Added: - Delete individual transcripts ## [0.8.354] - 2023-05-18 ### Changed: - Refactoring in ASR code on macOS - Refactoring in ASR code on iOS - Consolidate ASR code, maintained in macOS - Upgrade whisper.cpp to v1.4.2 ## [0.8.353] - 2023-05-18 ### Added: - Pass filter when navigating from habits to habit settings ## [0.8.352] - 2023-05-18 ### Changed: - Outbox item style - Upgrade dependencies ### Fixed: - Remove print_special output in transcription on macOS ## [0.8.351] - 2023-05-18 ### Added: - Speech recognition in any language supported by Whisper ## [0.8.350] - 2023-05-17 ### Added: - Config flag for task management - Conditionally show add task action icon ### Changed: - Tasks moved to separate tasks tab - Upgraded dependencies (major versions) - Use new SearchBar widget - Use new SearchBar widget on Habits page ### Fixed: - Navigation after recording audio ## [0.8.349] - 2023-05-16 ### Changed: - Journal tab renamed to Logbook ## [0.8.348] - 2023-05-14 ### Changed: - Remove `adaptive_dialog` library - Upgraded dependencies & removing unused - Upgraded Flutter to 3.10 - Upgraded Fluttium ### Fixed: - Flutter 3.10 warnings & errors - Android build issues after Flutter upgrade ## [0.8.347] - 2023-05-13 ### Added: - Settings page for downloading and activating whisper models - Download whisper.cpp models from Hugging Face - Detect downloaded models - Speech recognition on iOS - Background modes on iOS for fetch and processing - Select whisper model - Delete downloaded model ### Changed: - Only show speech settings on macOS and iOS for now - Hide large whisper models - Upgraded dependencies - Preparation for Flutter 3.10 - Only detect audio language when non-english model is selected ### Fixed: - Model download on iOS - Update model status after deleting model ## [0.8.346] - 2023-05-11 ### Added: - Automatically transcribe audio on macOS ## [0.8.345] - 2023-05-10 ### Added: - Transcription icon in audio player on macOS - Audio conversion from aac to wav with ffmpeg_kit_flutter - whisper.cpp library for speech recognition - English speech recognition, with the result logged - Transcript data structure - Add transcriptions to audio entries - Capture duration of generating transcript - Faster transcripts via compiler flag - Compiler flags -O3 -DNDEBUG - Update whisper.cpp to v1.4.0 - Compiler flags -O3 -DNDEBUG in debug/dev mode (transcription time for 1m test audio down from 27s to 4s) - Display audio duration in journal card - Language detection (logging only) - Indicator for existing transcriptions - Set entry text from transcript & update ### Changed: - Toggle for showing individual transcripts - Chore: upgraded dependencies ### Fixed: - Set entry text from transcript - Update entry text after adding transcript ## [0.8.344] - 2023-05-05 ### Changed: - Upgraded dependencies - Refactoring: BloC for state management on dashboards page ### Added: - Filter dashboards by categories - Tests for dashboards page ## [0.8.343] - 2023-05-04 ### Fixed: - Margin below habit completion dialog - Awaiting in getIt init ### Changed: - Upgraded dependencies ## [0.8.341] - 2023-05-02 ### Changed: - Upgraded dependencies - Show config flag for invalid cert - App documents folder on Windows ### Fixed: - App ID on Windows and Linux ### Added: - Save and restore window position on Windows and Linux - HotKey support on Windows and Linux ## [0.8.340] - 2023-04-29 ### Fixed: - Habit completion bottom sheet height ## [0.8.339] - 2023-04-28 ### Fixed: - Missing `libmpv.so` on Android preventing startup ## [0.8.338] - 2023-04-28 ### Fixed: - Screenshots of habit completion dialog in Fluttium flow ## [0.8.337] - 2023-04-28 ### Changed: - Improved habit completion bottom sheet - Upgraded dependencies ### Fixed: - Keyboard hiding habit completion bottom sheet ## [0.8.336] - 2023-04-28 ### Changed: - Smaller APK sizes (thanks @alexmercerind) - Upgraded dependencies - Faster CI runs on Buildkite ## [0.8.334] - 2023-04-27 ### Changed: - Dashboard in habit completion now above the dialog, leaving the dialog always in the same position ## [0.8.333] - 2023-04-26 ### Fixed: - Editor losing focus after interacting with the editor toolbar ## [0.8.333] - 2023-04-26 ### Changed: - Reduced APK size ## [0.8.332] - 2023-04-26 ### Fixed: - Location on Linux ## [0.8.331] - 2023-04-25 ### Added: - Audio playback on Linux ### Changed: - Upgraded dependencies ## [0.8.330] - 2023-04-24 ### Changed: - Sync assistant styling - Upgraded dependencies ### Fixed: - Sync getting stuck after generating new sync key and on reading sync message encrypted with old key ## [0.8.329] - 2023-04-23 ### Changed: - Upgrade flutter_quill lib ## [0.8.328] - 2023-04-23 ### Changed: - Show audio player inline in linked entries ## [0.8.327] - 2023-04-22 ### Fixed: - Navigation after delete (not when displayed as a linked entry) ### Changed: - Updated provisioning profile (was expired) - Automatically managed signing on iOS ## [0.8.326] - 2023-04-21 ### Added: - GitHub Action for running Fluttium tests on Windows ## [0.8.325] - 2023-04-20 ### Added: - Semantic labels in measurable data type setting - Screenshot of creating measurable data types for manual ### Changed: - Colors and whitespace - Upgraded dependencies - Initial height of chart selector bottom sheet - Slightly less saturated card color ### Fixed: - Test flows ## [0.8.324] - 2023-04-19 ### Fixed: - Image card background ## [0.8.323] - 2023-04-18 ### Changed: - Updated manual - Less verbose logging - App icon with gradient ## [0.8.322] - 2023-04-16 ### Changed: - Fluttium screenshots are pushed to lotti-docs repository ## [0.8.321] - 2023-04-15 ### Changed: - App icon on Android - App icon on macOS - App icon on Windows ## [0.8.320] - 2023-04-15 ### Changed: - App icon on Windows - Upgraded dependencies - Improved color picker ### Added: - Category color text field for HEX color, with ColorPicker moved to bottom sheet ### Fixed: - Update color HEX field after picking new color ## [0.8.319] - 2023-04-13 ### Changed: - Measurable setting page layout ## [0.8.318] - 2023-04-13 ### Changed: - New app icon ## [0.8.317] - 2023-04-12 ### Added: - Settings icon on Habits page ### Fixed: - SafeArea around Dashboards page ## [0.8.316] - 2023-04-12 ### Fixed: - Clear categories filter visibility - Text overflow for long habit titles ## [0.8.315] - 2023-04-12 ### Added: - Category selection for dashboards ### Changed: - Improved search field - Improved dashboards page header ## [0.8.314] - 2023-04-11 ### Changed: - Material icons in bottom nav - Material icons in audio recorder - Remove unused code - Upgraded dependencies ## [0.8.313] - 2023-04-11 ### Changed: - Updated README ## [0.8.312] - 2023-04-10 ### Changed: - Upgraded dependencies - Refactoring on Journal page - Improved habits page header ### Fixed: - SafeArea around Habit page ## [0.8.311] - 2023-04-09 ### Changed: - Initial window size on macOS ## [0.8.310] - 2023-04-09 ### Added: - Integration tests using Fluttium - Config flag for recording geolocation - Category color icons for habit completion card and entry detail view - Fluttium test in Buildkite ### Changed: - Reordered fields in habit config - Default story removed from habits ### Fixed: - Entry DateTime field width ## [0.8.309] - 2023-04-08 ### Changed: - Refactoring in audio recording ## [0.8.308] - 2023-04-07 ### Changed: - Data capture dialog style ## [0.8.307] - 2023-04-07 ### Changed: - Upgraded dependencies - App icon for windows msix - Removed chart animation - Whitespace in habit completion bottom sheet ## [0.8.306] - 2023-04-06 ### Added: - Windows build ## [0.8.305] - 2023-04-05 ### Changed: - Carousel in dashboards removed ### Fixed: - Data capture from dashboard line chart - Style consistency in dialog input fields ## [0.8.304] - 2023-04-05 ### Added: - Android release ## [0.8.303] - 2023-04-04 ### Changed: - Habit completion in modal bottom sheet instead of dialog - Measurement dialog without beamer page - Habit completion dialog without beamer page ### Added: - Select dashboard to display during habit completion - Show dashboard associated with habit in habit completion modal bottom sheet ### Fixed: - Sort & filter dashboards in habit definition - Complete habits for prior days ## [0.8.302] - 2023-04-03 ### Fixed: - Suggested previous values in measurement capture dialog cannot be selected ## [0.8.301] - 2023-04-03 ### Fixed: - Permission for notifications ### Changed: - Postponed request for geolocation in first-time user experience ## [0.8.300] - 2023-04-02 ### Fixed: - Picking multiple images only picking one image ## [0.8.299] - 2023-04-02 ### Changed: - Upgraded Flutter - Change habit active status to archived status ### Fixed: - DateTime form field width ## [0.8.298] - 2023-04-01 ### Changed: - Categories filter applies to all sections, not only currently due habits ## [0.8.297] - 2023-04-01 ### Changed: - Upgrade dependencies - Whitespace on entry detail page ### Fixed: - Styling of DateTime modal bottom sheet ## [0.8.296] - 2023-03-30 ### Fixed: - Habit completion chart display range issue ## [0.8.295] - 2023-03-29 ### Changed: - Upgraded dependencies - Improved naming in habit definition ### Fixed: - Habit Start Date field appeared filled when it was not ## [0.8.294] - 2023-03-28 ### Fixed: - Show save button upon category color change ## [0.8.293] - 2023-03-28 ### Changed: - Fill survey page removed ## [0.8.292] - 2023-03-27 ### Changed: - FormBuilderCupertinoDateTimePicker replaced in task estimate - FormBuilderCupertinoDateTimePicker replaced in habit completion dialog - Removed opacity in bottom sheets - FormBuilderCupertinoDateTimePicker replaced in new measurement dialog - New measurement page removed - flutter_datetime_picker removed from entry datetime modal - Simpler display of habit completion count - FormBuilderCupertinoDateTimePicker replaced in habit definition - FormBuilderCupertinoDateTimePicker code removed - Upgraded dependencies ## [0.8.291] - 2023-03-27 ### Fixed: - Missing day in habits after switching to DST ## [0.8.290] - 2023-03-26 ### Added: - Display time spent on task ## [0.8.289] - 2023-03-26 ### Fixed: - Category save duplicate warning - Disable category save icon when form invalid - Recreate category with the same name as previously deleted category - Scroll in setting when window is small ### Changed: - Upgraded dependencies ## [0.8.288] - 2023-03-25 ### Added: - Visualization for categories of open habits - Category name validation checking for duplicates - Categories filter bottom sheet with categories toggle - Filtered habits view by selected category ### Changed: - Upgraded dependencies ### Fixed: - Delete category question and confirmation label - Scroll in category bottom sheet - Prevent duplicate categories - Categories filter not visible when no open habits displayed - 180 days in habit completion rate chart ## [0.8.287] - 2023-03-22 ### Changed: - Habit completion card layout ## [0.8.286] - 2023-03-21 ### Added: - Category entity type - Categories list page - Categories details page - Set category color - Priority switch in habit, for more prominent display - Priority icon in habit settings list card - Priority icon in habits list card - Category selection in habit definition - Category color in habit settings card - Habits sorted by priority first - Habit color in habit completion card ### Changed: - Whitespace in settings - Upgraded dependencies ### Fixed: - Focus issue in habit category selection ## [0.8.285] - 2023-03-19 ### Added: - Segmented control for filtering which habits are shown (due, later today, complete, all) - Search field for habits ### Changed: - Selectable habit time spans - More obvious habit completion state with strike-trough text and subtle opacity - Upgraded dependencies - Toggle display of habits time span ## [0.8.284] - 2023-03-19 ### Changed: - Habit completion icon ### Fixed: - CI pipeline ## [0.8.283] - 2023-03-18 ### Changed: - Refactor: remove unused code - Line in header removed - Upgraded dependencies ## [0.8.282] - 2023-03-17 ### Added: - Placeholder text in editor (English and German) ### Changed: - Improved journal filters ## [0.8.281] - 2023-03-16 ### Changed: - Habit cards - Task status styling - Whitespace around cards - Task card whitespace - Refactoring: extracted theme ## [0.8.280] - 2023-03-16 ### Changed: - Use Material Design icons ## [0.8.279] - 2023-03-15 ### Changed: - Using Cards from Material Design 3 throughout where appropriate - Upgraded dependencies - Material Cards in Tag, Habit, and Dashboard definition pages ### Fixed: - Outbox color by status ## [0.8.278] - 2023-03-13 ### Changed: - Unused code removed - Upgraded dependencies ## [0.8.277] - 2023-03-03 ### Changed: - Upgraded dependencies ### Fixed: - Entry DateTime modal layout ## [0.8.276] - 2023-02-28 ### Changed: - more consistent bottom sheet modals - remove limit in tag search results - replace monospace font with main font & tabularFigures font feature ## [0.8.275] - 2023-02-26 ### Changed: - Refactor: use showModalBottomSheet for managing tags - CI: retry on exit status 2 - Upgrade fl_chart library ## [0.8.274] - 2023-02-26 ### Changed: - More consistent app bar - Reduce clutter in tasks header by moving count stats - Upgraded dependencies - Improved styling of about page ## [0.8.273] - 2023-02-24 ### Changed: - Upgraded dependencies - Progress bar height and opacity - Remove entry duration in journal card ## [0.8.272] - 2023-02-24 ### Changed: - Replace task progress indicator with material widget - Editor toolbar styling ## [0.8.271] - 2023-02-23 ### Fixed: - Background color at the top of the survey bottom sheet - Background color of survey dismiss dialog ## [0.8.270] - 2023-02-22 ### Changed: - Reduce clutter in entry card - Upgraded dependencies ## [0.8.269] - 2023-02-20 ### Changed: - Replace modal_bottom_sheet lib with Flutter's own implementation - Replace badges lib with Flutter's own implementation ## [0.8.268] - 2023-02-19 ### Fixed: - Habit fail button color - Save button color - Clipped save button on measurement page ## [0.8.267] - 2023-02-19 ### Changed: - Upgraded dependencies ## [0.8.266] - 2023-02-13 ### Changed: - Upgraded dependencies ### Fixed: - Data capture dialog jumpiness - Prevent dialog resize when save button becomes visible ## [0.8.265] - 2023-02-13 ### Fixed: - Navigate back icon on task page - Unordered list color in editor - Editor menu background color - Chip style - Primary material color ### Changed: - Improved task input fields layout - Improved chip layout - Tweak spacing in journal filters - Improved styling in settings - Upgraded dependencies - Tweak bottom navigation bar styling - Darker link and success icon color in habits completion ## [0.8.264] - 2023-02-11 ### Fixed: - Audio playback restart on navigate to audio entry ### Changed: - Improved input field layout ## [0.8.263] - 2023-02-10 ### Changed: - Refactoring: set main font in theme in one place - Material Design 3 enabled ## [0.8.262] - 2023-02-10 ### Fixed: - Display of null title when completing habit ### Changed: - Upgrade Flutter and dependencies ## [0.8.261] - 2023-02-09 ### Fixed: - Sleep data import - Flights of stairs data import - Total distance in interval data import - Jumpy badge animation ## [0.8.260] - 2023-02-08 ### Changed: - Upgraded very_good_analysis lib ## [0.8.259] - 2023-02-08 ### Changed: - Use Flutter 3.7.1 - Latest health lib (breaks flights of stairs and sleep types) ## [0.8.258] - 2023-02-07 ### Changed: - Replace read-only flutter_quill with flutter_markdown for better scroll performance ### Fixed: - Journal card text color when using bright theme ## [0.8.257] - 2023-02-06 ### Fixed: - Locking issue in sync ### Added: - Settings DB ### Changed: - Window manager persistence moved to settings database - Routing persistence moved to settings database - Last read UID persistence moved to settings database - Improved logging ## [0.8.255] - 2023-02-02 ### Changed: - Upgrade dependencies ### Fixed: - Keychain locking issue ## [0.8.254] - 2023-01-27 ### Changed: - Improved workout labels ### Fixed: - Remove broken workout time health chart (workouts are a separate category now) ## [0.8.253] - 2023-01-25 ### Changed: - Upgrade dependencies - major versions ## [0.8.252] - 2023-01-25 ### Fixed: - Repeat tapping of mic overwrote old audio file - Pop audio recorder as expected Added: - Pause icon in audio recorder functionality ## [0.8.251] - 2023-01-24 ### Changed: - Remove unused theme config widget - Hide config flag: allow_invalid_cert ## [0.8.250] - 2023-01-24 ### Changed: - Remove redundant config flag: enable_beamer_nav - Remove redundant config flag: listen_to_global_screenshot_hotkey - Remove redundant config flag: show_tasks_tab - Remove redundant config flag: hide_for_screenshot - Remove redundant config flag: notify_exceptions ## [0.8.249] - 2023-01-23 ### Changed: - Differentiate between failed and missing habit completion - Upgraded dependencies ## [0.8.248] - 2023-01-22 ### Changed: - Upgraded dependencies ### Fixed: - Fix updating habit completion type ## [0.8.247] - 2023-01-20 ### Changed: - Audio recording lib replaced ### Added: - Audio recording on macOS ## [0.8.246] - 2023-01-19 ### Changed: - Chore: redundant dependencies removed - Chore: upgraded dependencies ## [0.8.243] - 2023-01-18 ### Changed: - Adaptive minY value in habit completion chart - Toggle between zero-based and adaptive charts - Upgraded dependencies ### Fixed: - Habit completion percentages - Show toggle for adaptive charts only when leading to discernible differences ## [0.8.242] - 2023-01-16 ### Fixed: - AppStore Connect warning ## [0.8.241] - 2023-01-15 ### Fixed: - Record audio note as comment to other entry types ## [0.8.240] - 2023-01-14 ### Fixed: - Text color for entry duration when timer running - Fix possibility of accidentally overwriting task title, estimate, and status ## [0.8.239] - 2023-01-14 ### Changed: - Long press on entry type filter to select only one type - Select all entry types toggle button - Long press on task status filter to select only one status - Select all task statuses toggle button ### Fixed: - Add missing habit completion summary in entry detail view ## [0.8.238] - 2023-01-14 ### Fixed: - Code font color in text editor - Consolidate monospace text styles ## [0.8.237] - 2023-01-14 ### Fixed: - Navigate back after entry deletion ## [0.8.236] - 2023-01-14 ### Fixed: - Unlinking entries ## [0.8.235] - 2023-01-13 ### Added: - Remove habit streaks section - not useful, streaks are apparent without - Chore: upgraded dependencies ## [0.8.234] - 2023-01-13 ### Changed: - Journal card styling - Entry type filter styling - Task filter styling ## [0.8.233] - 2023-01-13 ### Changed: - Decluttered task view: only show editor toolbar on first focus - Decluttered task view: only show tags in task comments when not equal to parent tags - Show total time spent on a task - Show task stats - Task search header styling ## [0.8.232] - 2023-01-13 ### Changed: - Tasks by status can now be found in the journal tab - Full-text search in tasks - Tasks tab removed - Task card navigation adapted - CMD-R for reloading journal page - Flagged entries count badge move into search header - In-progress tasks count badge move into search header ### Fixed: - Assigning tags to tasks ## [0.8.231] - 2023-01-12 - Fix keyboard dismissal in search field by always showing the X icon - Fix clearing story selection in habit definition ## [0.8.230] - 2023-01-11 ### Changed: - Simplify settings by removing irrelevant favorite status switch in measurables (not used) - Story time chart selections removed, will need to be simplified and/or rethought - Dashboard notification time removed, will be handled better by notifications on habits, such as the habit of looking at a particular dashboard ## [0.8.229] - 2023-01-11 ### Fixed: - Update of JournalImage after change ## [0.8.229] - 2023-01-10 ### Changed: - Show popular values in capture dialog for past dates ## [0.8.228] - 2023-01-10 ### Added: - Full-text search database using FTS5 - Wire full-text database search (no refresh yet) - Add tags in full-text search - Add entities cache for faster lookup of measurable data types - Use entities cache in measurement summary - Refactor: move fetch logic to cubit - Refresh results when typing in full-text search field - Add new and updated text to full-text index - Fix index creation maintenance task - Remove previous entry in FTS5 index when updated - One-step index recreation ### Fixed: - Update JournalCard in infinite scroll automatically on change, e.g. after navigating back - Clear query ### Changed: - Upgraded dependencies ## [0.8.227] - 2023-01-07 ### Changed: - New search header in journal - Upgraded dependencies ## [0.8.226] - 2023-01-07 ### Changed: - Declutter task form - Simplified editor toolbar - Unified searchable list - Hide task title label when task defined - Floating search bar replaced in settings - Floating search bar replaced in journal ## [0.8.225] - 2023-01-06 ### Changed: - Improved text editor layout ## [0.8.224] - 2023-01-05 ### Fixed: - Disappearing keyboard on mobile - Color theme in tag search ## [0.8.223] - 2023-01-04 ### Changed: - Consistently use light and dark keyboard types - Upgraded dependencies ## [0.8.222] - 2022-12-31 ### Changed: - Limit editor height so that editor toolbar always stays visible ## [0.8.221] - 2022-12-30 ### Changed: - Fix completion rate by taking into account from when on to count a habit ## [0.8.220] - 2022-12-29 ### Fixed: - Screenshot delay - Reduce allocations in sync ## [0.8.219] - 2022-12-28 ### Changed: - Disable smooth curved lines in habits chart: sharp lines are not overshooting - Improved chart layout and CTA (tap chart for daily breakdown) ## [0.8.218] - 2022-12-27 ### Changed: - Journal page with infinite scroll ## [0.8.217] - 2022-12-26 ### Changed: - Improved blood pressure chart ## [0.8.216] - 2022-12-26 ### Changed: - Improved whitespace in habit page header - Habit chart grid: lines at 20, 40, 60, 80, 100% - Emphasized 80%-line (sensible minimum target) ## [0.8.215] - 2022-12-25 ### Changed: - Improved habit chart info on tap - Clear habit chart info within 15 seconds ## [0.8.214] - 2022-12-24 ### Added: - Stacked habit success chart, with success, skipped, explicitly failed, implicitly failed ## [0.8.213] - 2022-12-23 ### Added: - Throttle habits success scoring - Habit completions via click on habits success indicator ### Fixed: - Condition where sync inbox could fail during processing and be skipped - Performance issues when syncing health-related entries ## [0.8.212] - 2022-12-19 ### Changed: - Remove useless entry text toggle icon ## [0.8.211] - 2022-12-18 ### Changed: - Header position on image entries ### Fixed: - Duplicate display issue - Tag display issue in JournalCard ## [0.8.210] - 2022-12-17 ### Changed: - Improve line charts by using fl_chart library - Improve blood pressure chart by using fl_chart library ## [0.8.209] - 2022-12-16 ### Fixed: - Habit completion percentage when private habits not shown ## [0.8.208] - 2022-12-15 ### Added: - Charts for habit skip and explicit habit success ## [0.8.207] - 2022-12-13 ### Changed: - Unlink icon in editor toolbar instead of using Dismissable ## [0.8.206] - 2022-12-11 ### Added: - Tooltips for habit completions, showing date ### Changed: - Show habit streaks count at bottom of habits page, with labels - Extract habit streak lists & use adaptive header - Remove autofocus on measurement value field - Unify segmented time span controls on dashboard and habit pages - Increase analyzed habit completion time span to 90 days - Show date in habit completion rate chart tooltip ## [0.8.205] - 2022-12-09 ### Added: - Chart for habit completion rate over time ### Changed: - Improved whitespace in habit success indicators - Improved habit completion rate tooltips ## [0.8.204] - 2022-12-07 ### Added: - Progress bar for habit progress for the current day ## [0.8.203] - 2022-12-05 ### Added: - Suggest last used value in measurable - Upgrade dependencies - More responsiveness in measurement creation ## [0.8.202] - 2022-12-01 ### Added: - Habit autocomplete rules editor experiments ## [0.8.201] - 2022-11-22 ### Added: - Show habit description during completion ## [0.8.200] - 2022-11-21 ### Changed: - Refactoring in habit definition page: move logic to cubit ## [0.8.199] - 2022-11-20 ### Fixed: - Show expected habit success from when defined ## [0.8.198] - 2022-11-20 ### Added: - Editor display toggle ## [0.8.197] - 2022-11-20 ### Added: - Habit autocomplete definition refinements - Optional titles in habit rules - Habit autocompletion type other habit rule ### Changed: - New Flutter version & dependencies - Habits displayed for last 30 days by default on both mobile and desktop ## [0.8.196] - 2022-11-10 ### Changed: - Define default story for habit completion entry ## [0.8.195] - 2022-11-08 ### Changed: - Record duration between opening habit and completing habit (unless date in the past is selected) ## [0.8.194] - 2022-11-08 ### Changed: - Audio playback speed toggle instead of individual icon button ## [0.8.193] - 2022-11-07 ### Changed: - Upgraded dependencies - Error handling in editor ### Added: - CMD-S in task title field ## [0.8.192] - 2022-11-07 ### Changed: - Multiline comments and max width in habit capture dialog ## [0.8.191] - 2022-11-05 ### Changed: - Allow setting habit active or inactive - Rename "active from" to "expect success from" ## [0.8.190] - 2022-11-04 - Flutter upgrade - Upgraded dependencies ## [0.8.189] - 2022-10-30 ### Fixed: - Flicker in entry & task tag selection ## [0.8.188] - 2022-10-29 ### Fixed - Flickering keyboard issue when creating habit on mobile - Flickering keyboard issue when creating measurable data type on mobile - Flickering keyboard issue when creating dashboard data type on mobile ## [0.8.187] - 2022-10-29 ### Changed: - Flutter upgrade to 3.3.6 ## [0.8.186] - 2022-10-29 ### Changed: - Improved habit completion add icon ## [0.8.185] - 2022-10-28 ### Changed: - Simplify sync outbox, remove faulty network connected check - Add sync inbox tests for decrypting and writing image and audio files ## [0.8.184] - 2022-10-28 ### Fixed: - Update habits range after midnight ## [0.8.183] - 2022-10-27 ### Added: - Autocomplete habits data structure - Skipping habit doesn't break the chain ## [0.8.182] - 2022-10-24 ### Changed: - Screenshot exception logging - Retry IMAP actions with exponential backoff ## [0.8.181] - 2022-10-23 ### Changed: - Styling: entry icons - Styling: colors ## [0.8.180] - 2022-10-23 ### Changed: - Move Sync inbox to separate isolate ## [0.8.179] - 2022-10-22 ### Changed: - Simplify sync by reusing IMAP client in one place - Restart outbox client isolate on network reconnect ## [0.8.178] - 2022-10-21 ### Changed: - Upgraded dependencies - Count habits total and finished today - Count habit streaks of three days (up until yesterday) - Count habit streaks of one week (up until yesterday) - Sections for longer and shorter streaks ## [0.8.177] - 2022-10-17 ### Fixed: - Bring back index creation in journal database - Fix habit success indicator width ## [0.8.176] - 2022-10-16 ### Fixed: - "Task not found" when task still loading - Spacing between habit success indicators ## [0.8.175] - 2022-10-16 ### Added: - Habit completion types success, skip, fail ## [0.8.174] - 2022-10-16 ### Changed: - Removed index creation in JournalDb for now ## [0.8.173] - 2022-10-16 ### Changed: - Improved photo view ## [0.8.172] - 2022-10-16 ### Changed: - Upgraded dependencies - JournalDb moved to a separate isolate, freeing up CPU resources on the main thread/isolate ## [0.8.171] - 2022-10-13 ### Changed: - Sort habits by show from time, then a-z ### Fixed: - Time chart didn't include today ## [0.8.170] - 2022-10-12 ### Changed: - Enable isolate support in JournalDb, SyncDb, LoggingDb - Run SyncDb and LoggingDb in separate isolate (thread) - Run Sync outbox in isolate to avoid jank ## [0.8.169] - 2022-10-11 ### Changed: - Upgrade Flutter to 3.3.4 - Upgrade dependencies ## [0.8.168] - 2022-10-10 ### Changed: - Entry details header icons ## [0.8.167] - 2022-10-08 ### Changed: - Delayed health import to improve scroll performance ## [0.8.166] - 2022-10-06 ### Added: - Health import for DISTANCE_WALKING_RUNNING ## [0.8.165] - 2022-10-06 ### Changed: - Increase measurement line chart height ### Fixed: - Alignment of time axis between different types ## [0.8.164] - 2022-10-04 ### Added - Habits definition in Settings - Add habit chart in dashboard - Habits tab - Habits grouped by open/completed - Localize open/closed headers - Add show from field ## [0.8.163] - 2022-10-01 ### Changed: - Audio recorder icons in dark mode ## [0.8.162] - 2022-10-01 ### Changed: - Active icons in surveys ## [0.8.161] - 2022-10-01 ### Changed: - Launch background color ## [0.8.160] - 2022-09-30 ### Fixed: - Sleep import on iOS ## [0.8.159] - 2022-09-29 ### Added: - Aggregation by hour in measurables charts ## [0.8.158] - 2022-09-29 ### Changed: - Colors adapted for dark mode - VU meter in audio recording indicator removed ## [0.8.157] - 2022-09-29 ### Changed: - Segmented control for dashboard time span - Style: dark mode - Style: improved segmented control for dashboard time span ## [0.8.156] - 2022-09-26 ### Changed: - Charts hidden in journal cards - New bottom navigation - Measurement capture dialog style - AppBar style with leading text - Splash screen color - Improved add icons in tasks, dashboards, measurables - Hide audio recording indicator when on recorder page ### Fixed: - Chart header cut off ## [0.8.155] - 2022-09-25 ### Changed: - Improve filling surveys by using modal_bottom_sheet lib - Use showCupertinoModalBottomSheet instead of showModalBottomSheet ## [0.8.153] - 2022-09-25 ### Fixed: - Story assignment on mobile ## [0.8.152] - 2022-09-24 ### Changed: - DefinitionCard design - Design: Settings > Tags - Design: Settings > Dashboard Management - Design: Settings > Measurables - Consolidated settings cards ### Fixed: - Health chart legend on select ## [0.8.151] - 2022-09-22 ### Changed: - Style: Inconsolata as monospace font - New color theme in journal and tasks list - New color theme in entry details ## [0.8.150] - 2022-09-20 ### Fix: - Bar overlapping domain axis ### Changed: - Design tweaks in measurement capture - Design tweaks survey capture - Replace Lato font - Empty dashboards page with how to use button - Bar chart style ## [0.8.149] - 2022-09-18 ### Added: - Detect when desktop app is resumed in Sync ## [0.8.148] - 2022-09-18 ### Changed: - Dashboard chart header tweaks - No hover color on IconButton elements on desktop - Chart colors ## [0.8.147] - 2022-09-18 ### Changed: - Sync Conflicts layout - Sync Conflicts resolution UI ## [0.8.146] - 2022-09-17 ### Changed: - Style: Icon alignment in Settings > Advanced - Debug logging for Sync - Style: barrier color in new measurement modal - Style: add measurement icon - Style: floating action button color - Style: white app bar in dashboards - Style: app bar redesign - Style: remaining charts ## [0.8.145] - 2022-09-15 ### Changed: - Styling: Settings layout - Styling: hover in Settings - Styling: measurables chart - Styling: survey chart - Styling: health chart - Styling: BP chart - Styling: BMI chart ## [0.8.144] - 2022-09-13 ### Changed: - Move version and entry count to about page ## [0.8.143] - 2022-09-13 ### Changed: - Typography: use PlusJakartaSans in Settings - Typography: use PlusJakartaSans in Settings > Tags - Typography: use PlusJakartaSans in Settings > Dashboards - Typography: use PlusJakartaSans in Settings > Advanced - Typography: use PlusJakartaSans in Settings > Advanced > Maintenance - Typography: use PlusJakartaSans in Settings > Config Flags - Typography: use PlusJakartaSans in app bar - Typography: use PlusJakartaSans in bottom navigation - Hide icons in Settings - White background in Settings - White background in Dashboards List - White background in Dashboards - Chart title in black - Move app version to Settings > About ## [0.8.142] - 2022-09-13 ### Changed: - Upgraded dependencies - Assign story tag to comment entries as well ### Added: - More tests for persistence logic ## [0.8.141] - 2022-09-12 ### Changed: - Flutter version upgrade - Upgraded dependencies - Sync reliability improvements ## [0.8.140] - 2022-09-08 ### Fixed: - Bottom sheet for tag selection overlaying the bottom nav bar ## [0.8.139] - 2022-09-07 ### Changed: - Upgrade dependencies ## [0.8.138] - 2022-09-06 ### Added: - Tests for surveys ## [0.8.137] - 2022-09-06 ### Changed: - Navigation: tap on open tab navigates to tab root ## [0.8.136] - 2022-09-05 ### Changed: - Upgraded dependencies - Added tests for Audio Player widget - Added tests for Audio Recorder widget ## [0.8.134] - 2022-09-02 ### Changed: - Upgraded dependencies - Capture text with adding a measurement ## [0.8.133] - 2022-08-29 ### Changed: - Allow adding text in measurable entries - Allow adding text in survey entries ## [0.8.132] - 2022-08-26 ### Changed: - Measurements captured in alert dialog, not modal - Cross-tab navigation - Navigate to dashboard from settings ## [0.8.131] - 2022-08-24 ### Changed: - Plus icon for adding measurement from dashboard instead of double tab - Plus icon for filling survey from dashboard instead of double tab ## [0.8.130] - 2022-08-23 ### Added: - New navigation using beamer (in progress, available via config flag) - Navigation in Settings > Tags using beamer - Navigation in Settings > Dashboards using beamer - Navigation in Settings > Measurables using beamer - Navigation in Settings > Config Flags using beamer - Navigation in Settings > Health Import using beamer - Navigation in Settings > Advanced using beamer - Navigation in Journal using beamer - Navigation in Tasks using beamer - Navigation in Dashboards using beamer ## [0.8.129] - 2022-08-17 ### Changed: - Copy SyncConfig to clipboard, encrypted with random one-time password - Paste encrypted SyncConfig from clipboard & decrypt with one-time password ## [0.8.128] - 2022-08-11 ### Fixed: - Text wrap in config flags on small screen (e.g. iPhone 12 mini) ## [0.8.127] - 2022-08-11 ### Changed: - Navbar changed to Salomon style ### Fixed: - Navigation glitch where the bottom nav bar was moving - Error handling when page does not exist ## [0.8.126] - 2022-08-08 ### Changed: - Updated dependencies - Inline code style in editor ## [0.8.124] - 2022-08-08 ### Changed: - Move dashboards page to left - Change dashboards header - Whitespace tweaks ## [0.8.124] - 2022-08-06 ### Changed: - Add toggle icon button for map visibility in entry header - Remove map toggle in entry footer ## [0.8.123] - 2022-08-05 ### Changed: - Save running timer progress ## [0.8.122] - 2022-08-02 ### Changed: - Entry details layout ## [0.8.120] - 2022-08-02 ### Fixed - Crash in tags modal due to wrong context ## [0.8.120] - 2022-08-01 ### Changed: - Time record icon only on text entries ## [0.8.119] - 2022-07-30 ### Changed: - Decoupling of PersistenceLogic and widgets - Decoupling of JournalDb and widgets ## [0.8.117] - 2022-07-28 ### Changed: - Show unsaved state after changing task title - Save entries when navigating away - Number format for selected measurement in chart - Only show dashboard slideshow icon when multiple dashboards are defined - Display more obvious entry save button - Spacing in entry header ## [0.8.116] - 2022-07-26 ### Fix: - Build errors ## [0.8.115] - 2022-07-25 ### Fixed: - Sync resetting its own offset - Polling ## [0.8.114] - 2022-07-21 ### Added: - Slideshow for dashboards ## [0.8.113] - 2022-07-20 ### Added: - Count duration for entries spanning multiple days for each individual day - Weekly aggregation in story time charts ## [0.8.112] - 2022-07-19 ### Changed: - Improved query for substring matched stories ## [0.8.111] - 2022-07-17 ### Changed: - Time format hh:mm:ss in time charts for aggregate of selected day - Time format hh:mm:ss in workout charts for aggregate of selected day ## [0.8.110] - 2022-07-15 ### Added: - Logging in sync ## [0.8.109] - 2022-07-15 ### Changed: - Improved whitespace in DateTime modal on mobile ## [0.8.108] - 2022-07-15 ### Added: - Wildcard matches in story charts ## [0.8.107] - 2022-07-14 ### Fixed: - Close photo button closes fullscreen photo ### Changed: - Better close photo icon, in white with black shadow ## [0.8.106] - 2022-07-14 ### Fixed: - Duration display as absolute value ## [0.8.105] - 2022-07-14 ### Added: - Confirmation dialog when unlinking entry ## [0.8.104] - 2022-07-13 ### Changed: - Improved logging in sync ## [0.8.103] - 2022-07-13 ### Changed: - Improved time chart ## [0.8.102] - 2022-07-12 ### Changed: - Simplified & cleaner Inbox Service ## [0.8.101] - 2022-07-11 ### Changed: - Show duration ## [0.8.100] - 2022-07-10 ### Changed: - Show single dashboard directly without dashboards list page ## [0.8.99] - 2022-07-10 ### Changed: - Show bottom navigation bar in dashboard page ## [0.8.98] - 2022-07-10 ### Fixed: - Dashboard save button not appearing after reordering items ## [0.8.97] - 2022-07-10 ### Changed: - Simplified & cleaner Outbox Service ## [0.8.96] - 2022-07-08 ### Changed: - Longer IMAP timeouts for better support of flaky connections - Fix toggle outbound sync in outbox monitor - Config flags for enabling inbox and outbox - Config flag for allowing invalid certificate (useful for testing, e.g. with [Toxiproxy](https://github.com/Shopify/toxiproxy)) ### Added: - Tests for measurement in journal - Tests for health entry in journal - Tests for dashboards measurement charts - Tests for dashboards health charts - Tests for dashboards workout charts - Tests for logging page ## [0.8.95] - 2022-07-07 ### Changed: - Refactoring (no UI changes) ## [0.8.94] - 2022-07-07 ### Changed: - Upgraded dependencies ## [0.8.93] - 2022-07-06 ### Added: - Tests for journal page - Tests for tasks page ## [0.8.92] - 2022-07-06 ### Added: - Tests for database and persistence logic ## [0.8.91] - 2022-07-05 ### Added: - Persistence of themes as JSON ## [0.8.90] - 2022-07-04 ### Changed: -Dependencies (no UI changes) ## [0.8.89] - 2022-07-04 ### Added: - Refactor theme management for color picker - Config flag for color picker on desktop - Basic theming config with color pickers on desktop - Show previews and tap to expand/show picker in theme config - Toggle theme config display via menu ## [0.8.88] - 2022-07-01 ### Fixed: - Hex color strings now parsed like CSS colors ## [0.8.87] - 2022-07-01 ### Fixed: - Timezone for notification ## [0.8.86] - 2022-07-01 ### Added: - Support for inline code in editor - Support for strikethrough inline style in editor ## [0.8.85] - 2022-06-30 ### Changed: - FadeIn animation on new measurement page ### Fixed: - Navigation pop after changing entry date - Navigation pop after adding measurement - Entry text color in for creating measurable ## [0.8.83] - 2022-06-30 ### Fixed: - Remove notifications for deleted dashboards ## [0.8.82] - 2022-06-29 ### Added: - Bright ☀️ color scheme with config flag - Change️ color scheme from menu on desktop - Add loading screen for dashboards, with animation ## [0.8.80] - 2022-06-27 ### Changed: - Use AppRouter mock in tests (no UI changes) ## [0.8.79] - 2022-06-27 ### Changed: - Dependency injection for SecureStorage (no UI changes) ## [0.8.78] - 2022-06-25 ### Changed: - Improve dashboards search width on desktop ## [0.8.78] - 2022-06-24 ### Changed: - Theme colors - Whitespace in task card ## [0.8.77] - 2022-06-24 ### Added: - Tests for Settings (no UI changes) ## [0.8.76] - 2022-06-24 ### Changed: - Apple developer account for debug mode, for reasons ( ͡ಠ ʖ̯ ͡ಠ) - Should not affect the build pipeline, current version simply tests the pipeline ## [0.8.75] - 2022-06-24 ### Added: - Tests for measurables detail page ## [0.8.74] - 2022-06-24 ### Added: - Faster measurement entry on desktop with autofocus and Cmd-S hotkey - Widget tests for new measurement page (no UI changes) ## [0.8.73] - 2022-06-23 ### Added: - Widget tests for dashboard definition page (no UI changes) ## [0.8.72] - 2022-06-23 ### Added: - Link from new measurement page to respective measurable definition ## [0.8.71] - 2022-06-23 ### Fixed: - Line wrap for long title and description in dashboard definition card - Line wrap for long description in dashboard page - Line wrap for long title and description in measurement card - Line wrap for long title in dashboard chart header ## [0.8.70] - 2022-06-23 ### Changed: - Show private status in dashboard card - Show daily review time in dashboard card ## [0.8.69] - 2022-06-23 ### Changed: - Show unit name in measurement type card - Show aggregation type in measurement type card ## [0.8.68] - 2022-06-23 ### Changed: - Leading insights icon in measurement card removed ## [0.8.67] - 2022-06-23 ### Changed: - Outbox monitor layout - Outbox badge now displays larger counts ## [0.8.66] - 2022-06-21 ### Added: - Tests for Sync assistant widgets (no UI changes) - Tests for OutboxCubit (no UI changes) ## [0.8.65] - 2022-06-20 ### Changed: - Remove aggregation label in chart when aggregation none - Use aggregation none as default ## [0.8.64] - 2022-06-19 ### Added: - Tests for Sync assistant logic (no UI changes) - Tests for Sync assistant widgets (no UI changes) ## [0.8.62] - 2022-06-18 ### Changed: - Guard Save button in new measurement by validation ## [0.8.61] - 2022-06-17 ### Fixed: - Saving tags and other form data ## [0.8.60] - 2022-06-17 ### Changed: - Fill survey directly from dashboard ## [0.8.59] - 2022-06-17 ### Fixed: - Save dashboard without daily review time filled out ## [0.8.58] - 2022-06-17 ### Added: - Workout type swimming ## [0.8.57] - 2022-06-16 ### Fixed: - Grey boxes in flagged entries that do not have text yet - Header margin on mobile ## [0.8.56] - 2022-06-16 ### Fixed: - App bar when creating new entries ## [0.8.55] - 2022-06-16 ### Changed: - Improvements in Sync Assistant - Prevent progression in Sync Assistant when not allowed ## [0.8.54] - 2022-06-14 ### Added: - Check valid mail account in sync assistant - Check saved IMAP config in sync assistant ## [0.8.53] - 2022-06-13 ### Added: - App bar with save button in new measurement page - Ignore chart interaction on journal card ## [0.8.52] - 2022-06-13 ### Changed: - Dev playground removed, not useful ## [0.8.51] - 2022-06-13 ### Added: - Search field for tasks in full width - Search field for measurables in full width ## [0.8.50] - 2022-06-13 ### Changed: - Indicate unsaved changes on tag edit page - Indicate unsaved changes on measurable data type edit page - Indicate unsaved changes on dashboard edit page ## [0.8.49] - 2022-06-13 ### Fixed: - Timezone name on Linux ## [0.8.48] - 2022-06-13 ### Fixed: - Location on Linux ## [0.8.47] - 2022-06-12 ### Changed: - Improve first-time user experience for measurables ## [0.8.46] - 2022-06-12 ### Changed: - Improve layout of health data entry - Improve layout of measurable data entry ## [0.8.45] - 2022-06-12 ### Fixed: - Sync of entities without vector clock - Dashboards sorted alphabetically ### Changed: - Optional description field in dashboard definitions - Optional description and unit fields in measurable definitions ### Added: - Maintenance task for reprocessing messages ## [0.8.44] - 2022-06-12 ### Changed: - Layout improvements in empty dashboards page ## [0.8.43] - 2022-06-12 ### Changed: - Allow dashboards with the same name - Add maintenance task for purging deleted items ## [0.8.42] - 2022-06-11 ### Changed: - Auto-sizing text in Sync Assistant ## [0.8.41] - 2022-06-10 ### Changed: - Default IMAP folder for sync ## [0.8.40] - 2022-06-09 ### Fixed: - No health import on desktop ## [0.8.39] - 2022-06-09 ### Changed: - Ignore foreign messages in IMAP folder ## [0.8.38] - 2022-06-09 ### Fixed: - Romanian language support in forms ## [0.8.37] - 2022-06-09 ### Changed - Improved icons ## [0.8.36] - 2022-06-08 ### Added: - Screenshot from desktop menu ## [0.8.35] - 2022-06-08 ### Changed: - Save screenshots on desktop as JPG - Improved icons ## [0.8.34] - 2022-06-08 ### Added: - French translation ### Changed: - Improved Sync Assistant - Trim fields in email config - Label for Sync enable/disable ## [0.8.32] - 2022-06-07 ### Added: - Tooltips for circular add actions - Fix hover UX ## [0.8.32] - 2022-06-07 ### Fixed: - Navigation issue ## [0.8.31] - 2022-06-04 ### Changed: - Define aggregation type for dashboard item ## [0.8.29] - 2022-06-02 ### Added: - Empty Dashboards instructions ### Changed: - Delete dashboards confirmation in red - Save and View button in dashboard definition removed ## [0.8.29] - 2022-06-01 ### Added: - Beginnings of a Manual ### Removed: - Default measurable types ### Added: - AppBars with matching titles for dashboard and measurable data type management ## [0.8.27] - 2022-06-01 ### Changed: - Improved UI in header ## [0.8.26] - 2022-05-31 ### Fixed: - Top margin for iPhone notch ## [0.8.25] - 2022-05-31 ### Changed: - Hide tasks tab unless config flag is set ## [0.8.24] - 2022-05-31 ### Changed: - Improve Search Header in Tasks ## [0.8.23] - 2022-05-31 ### Changed: - Improve Search Header in Journal ## [0.8.22] - 2022-05-31 ### Changed: - Show open, groomed & in-progress tasks by default - Remove AppBar in Journal - Show all types in Journal by default ## [0.8.20] - 2022-05-28 ### Fixed: - Bring back workout import ## [0.8.20] - 2022-05-28 ### Fixed: - Editor crashes ## [0.8.19] - 2022-05-25 ### Added: - Romanian localization ## [0.8.17] - 2022-05-22 ### Fixed: - Intra-day steps import ## [0.8.16] - 2022-05-22 ### Added: - Dashboard not found header, this becomes relevant after deleting a dashboard ## [0.8.15] - 2022-05-21 ### Changed: - Activity imports up to now ## [0.8.14] - 2022-05-21 ### Fixed: - Dashboard creation ## [0.8.12] - 2022-05-20 ### Changed: - Disable file sharing on iOS ## [0.8.11] - 2022-05-19 ### Added: - Persistence of editor drafts ## [0.8.9] - 2022-05-18 ### Changed: - Add task header ## [0.8.7] - 2022-05-17 ### Changed: - Add task stats header for task list ## [0.8.5] - 2022-05-16 ### Changed: - GitHub release for Linux from GitHub Action ## [0.8.4] - 2022-05-16 ### Changed: - Faster screenshots when no location in cache ## [0.8.3] - 2022-05-16 ### Changed: - Bottom Navigation Bar hidden in entry details - Bottom Navigation Bar hidden in sync assistant - Bottom Navigation Bar hidden in logging ## [0.8.2] - 2022-05-16 ### Changed: - Bottom Navigation Bar hidden in dashboards - Dashboard title moved to app bar ## [0.8.1] - 2022-05-15 ### Changed: - Upgrade to Flutter 3.0.0 ## [0.7.38] - 2022-05-12 ### Added: - Min and Max weight in BMI chart range - Only allow charts to shift back, not forward ## [0.7.37] - 2022-05-11 ### Added: - Disable panning while zoom is ongoing ## [0.7.36] - 2022-05-11 ### Added: - Horizontal Chart Zoom - Horizontal Chart Panning ## [0.7.33] - 2022-05-11 ### Added: - Hide inactive dashboards ## [0.7.32] - 2022-05-11 ### Added: - Inherit private status from linked ## [0.7.31] - 2022-05-11 ### Added: - Dark keyboard on iOS ## [0.7.30] - 2022-05-11 ### Fixed: - Zero wait for location on entry create, will be added later when available ## [0.7.29] - 2022-05-10 ### Fixed: - Tabs routes are now restored on application restart ## [0.7.28] - 2022-05-05 ### Fixed: - Audio playback for multiple recordings in list of linked entries was not working previously ## [0.7.22] - 2022-04-28 ### Added: - Haptic feedback for setting/unsetting starred/private/flagged statuses ## [0.7.19] - 2022-04-28 ### Fixed: - Footer spacing on mobile ## [0.7.18] - 2022-04-27 ### Added: - Share image and audio files from share button in entry footer ### Changed: - Removed audio file sharing from audio player ## [0.7.17] - 2022-04-27 ### Added: - Share audio recordings ## [0.7.16] - 2022-04-27 ### Added: - No autofocus on task text ## [0.7.15] - 2022-04-27 ### Added: - Adaptive max height for images ## [0.7.14] - 2022-04-27 ### Fixed: - Code signing on macOS leading to crash on startup ## [0.7.13] - 2022-04-26 ### Fixed: - No sync calls when not configured ## [0.7.11] - 2022-04-26 ### Fixed: - Linux entry persistence crash fix ## [0.7.10] - 2022-04-26 ### Added: - Sync settings: hide sensitive info - Sync settings: select IMAP folder - Styling: shadow on navigation bar - Sync assistant ### Changed: - Task form styling - Unfocus on save only on mobile - Sync IMAP messages marked seen - Sync IMAP messages in lotti_sync folder - Styling: entry card color ## [0.7.9] - 2022-04-25 ### Changed: - Text color in sync settings - Unfocus on save - Entry styling ## [0.7.8] - 2022-04-24 ### Added: - New color scheme - Text editor in slideshow for faster import ## [0.7.7] - 2022-04-24 ### Added: - Added flutter_image_slideshow ## [0.7.6] - 2022-04-22 ### Added: - Release to TestFlight via fastlane - Added fastlane-plugin-changelog - Populating release notes/what to test from CHANGELOG - Added fastlane match - Added GitHub Actions setup ## [0.7.5] - 2022-04-21 ### Added: - Added `make` tasks for TestFlight upload ### UI/UX - Checklist section polish: - Clear visual separation for items with subtle rounded backgrounds and hairline borders - Increased row padding (H 16, V 8), multi‑line support (up to 4 lines) with fade overflow - Checked items use strikethrough with reduced text opacity for faster scanning - Swipe‑to‑delete background now clips to the same radius and aligns with row padding - Header keeps Edit and Export; Delete remains available while editing just below the header to ensure a reliable tap‑target (follow‑up planned to unify in header) - No functional changes to reordering, export/share, or AI suggestions ### Tests - Added widget tests for checklist visuals (strikethrough on check, row wrapper presence) and suggestion overlay rendering; kept behavioural tests green.