Each screen's button kept its own open state and its own draft, so the
question half written on the Library was gone on Home. ComposerCubit,
provided beside ConversationDockCubit, now holds whether the field is
open and the draft (text, files attached, parts chosen) for every
button. Switching tabs keeps it as it is (AppShell says so before
goBranch); going to another screen within a tab, or over the tabs, keeps
it open only while something is written or attached. Only the button of
the screen at the router's location unfolds, so no tab kept alive
behind holds a second field.
Whats-New: The question you are writing stays open as you switch tabs.
Whats-New-De: Die Frage, die du schreibst, bleibt offen, wenn du den Tab wechselst.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
The developer's review of 4 Nov 2026. Below the question, × and + stand
at the start, who answers is centred and the conversations and the
microphone stand at the end, every button the same rounded square, the
microphone listening and Send too. The row above the text ("Only what I
choose · 2 parts") goes, as +'s sheet offers the same choice; a web
search gets + with that row alone. Who answers shows its model's name
alone ('Haiku 4.5', AssistantProvider.modelName: Anthropic's without
"Claude", others as they are), outlined in its provider's colour (D60).
The microphone is halkyon_app's Dictation: what is said goes where the
cursor stands, and it listens again after each pause until stopped. It
gives way to Send once anything could be sent, as Claude's field does.
Whats-New: The question's buttons are in one row: close and + on the left, who answers in the middle, Conversations and the microphone on the right.
Whats-New: What goes along is chosen under +, for the web search too.
Whats-New: Who answers shows the model's name, in its provider's colour.
Whats-New: The microphone keeps listening after a pause, writing where the cursor is, until you stop it.
Whats-New-De: Die Knöpfe der Frage stehen in einer Reihe: Schließen und + links, wer antwortet in der Mitte, Unterhaltungen und Mikrofon rechts.
Whats-New-De: Was mitgeht, wählst du unter +, auch für die Websuche.
Whats-New-De: Wer antwortet, zeigt den Namen des Modells, in der Farbe seines Anbieters.
Whats-New-De: Das Mikrofon hört nach einer Pause weiter zu und schreibt, wo der Cursor steht, bis du es stoppst.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
On GrapheneOS, which installs no recogniser, the microphone did nothing
and stayed listening: speech_to_text's initialize throws a
PlatformException (recognizerNotAvailable) there rather than answering
false, and DeviceSpeechInput let it through. A listen now answers how it
went (SpeechStart: listening, noRecognizer, unavailable) and never
throws; Learn's field and Meditate's dream say that no speech
recognition is installed and name two to install (FUTO Voice Input,
Google's speech services), in halkyon_app's texts. It also passes
androidNoBluetooth, as the apps declare no Bluetooth permission the
plugin would otherwise ask for.
Whats-New: Where the device has no speech recognition, the microphone says so and how to get one.
Whats-New-De: Wo das Gerät keine Spracherkennung hat, sagt das Mikrofon das und wie du eine bekommst.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
The developer's review of 4 Nov 2026: "an actual conversation with the AI
chatbot is not desirable and necessary". The field's voice-mode button,
the screen it opened, reading answers aloud and its texts go, and with
them flutter_tts, its patched copy in third_party/ (one override fewer),
the TTS_SERVICE query in the manifest and the plugin's registrants. The
microphone stays and dictates into the question. ImageApproval is the
conversation sheet's own again, as only voice mode needed it.
Whats-New: Voice mode is gone; the microphone dictates into your question.
Whats-New-De: Der Sprachmodus ist weg; das Mikrofon diktiert in deine Frage.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
Beside the field's microphone, a button opens voice mode over the whole
screen: a glow rising from below, a big microphone, and +, who answers,
the speaker and End in one row. It listens with the device's recogniser,
asks what was heard in one kept conversation, reads the answer aloud
with the device's speech (flutter_tts, its Markdown left out) and
listens again. On Linux there is none, as there is no microphone; where
nothing reads aloud, answers are shown as text.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
Following the developer's review of 3 Nov 2026 (ADR 0021, D67):
- + below the assistant's field, and below a conversation's follow-up,
opens "Add to your question" as Claude's "Add to session": three big
tiles side by side, Camera (where the device takes photos) and Photos
for an assistant reading images, and Files; below them, in the field, a
row leading to what goes along, showing how far the assistant looks and
how many parts are chosen.
- Camera and Photos through image_picker (flutter.dev's own): the camera's
own app and the system's photo picker, no permission asked on Android, a
file dialog on the desktop, which has no Camera; photos made at most
2048 pixels a side, read on the device like any file attached.
- What goes along is chosen in rounded rows, as Claude's Permission, in
place of radio buttons and chips: how far it looks with a tick, private
studysets with a switch, the parts each with a tick, a card's image with
its thumbnail.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
Following the developer's review of 3 Nov 2026 (ADR 0021, D67):
- Unfolded in the Library, the field spans the screen between its margins,
as on Home and a studyset: it fell short on the right, by the padding
around the two folded buttons, which now stands only around those.
- Who answers stands right before the buttons at the right, which no longer
shift as its name changes (Claude Haiku, Claude Opus): it fills the space
left, aligned to its end, rather than a share of it.
- Every button of the field and of a conversation's follow-up is boxed, a
tile a shade above the field with rounded corners, as Claude's apps have
them; who answers too.
- On a studyset with a licence, the unfolded field takes its usual place
over the licence's line rather than squeezing in above it; folded, its
row still keeps clear of the line.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
Learn's `SpeechInput` and `DeviceSpeechInput` (ADR 0021, D67) are
halkyon_app's now, and its `FakeSpeechInput` is in `testing.dart`, so
Meditate's dream journal (ADR 0038, D15) listens through the same code.
Learn imports them from there and behaves as before: its composer and
its tests are unchanged but for the imports, and `DeviceSpeechInput()`
asks for what it asked before.
`DeviceSpeechInput` takes three options, off by default:
- `preferOnDevice`: on Android, recognition on the device where it can
(Android 12 on), else offline; should that fail before a word is
heard (not a pause), it listens as Android otherwise does from then
on.
- `pauseFor`: how long a pause may last before the recogniser stops.
- `listenFor`: how long one listen may last at most.
`Dictation` writes what is heard where the cursor stands
(`DictationInsertion`: in place of a selection, set apart by a space),
shows the words so far, and listens again after each pause while it is
on, stopping by itself after three listens running heard nothing.
speech_to_text moves from Learn's pubspec to halkyon_app's, the lock
unchanged. Todo, Train and Meditate link the plugin now; their
generated plugin registrants (macOS, Windows) say so. An app declares
the microphone itself, and Todo and Train declare none, so nothing
there can listen.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
The public profile was already one per person for every service of an
instance (the register, ADR 0016); its app side was Learn's alone. It
moves into halkyon_app: the model, the repository on the module's
endpoints, the form, the avatar drawn on the device and the sheet with
someone's profile. They take the Public prefix (PublicProfileRepository,
BasePublicProfileGateway), as Train's own ProfileRepository holds the
body's figures. AppDependencies gains profiles and files, the picker the
avatar is chosen with, which moves out of Learn too.
HalkyonProfile is every app's Profile screen: the person's own tab first,
the same in every app (the account and the ways they sign in, the app's
own cards, then the public profile), then the app's own tabs, folding into
the title. Learn adds its inbox and Statistics. Train adds Progress, Days
and, where nothing records, Plans, and so opens on the person's tab, no
longer on Progress. Todo, which has no Profile tab, opens it from a row at
the top of Browse, where Todoist keeps the account.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
The question's text stands above its buttons: + attaching files, who
answers as a chip (Effort now a submenu of it), the conversations kept,
and the microphone while nothing is written, which gives way to the
button sending. What goes along is one row above the text, opening a
sheet with how far the assistant looks and the parts to send. A
conversation's follow-up has the same + and microphone.
Files are read on the device: an image goes as itself to an assistant
that reads images, a PDF as its pages' text, a text file as its text,
each cut at 20,000 characters and labelled by its name after the
question.
Voice is the device's own speech recognition (speech_to_text: Android,
Apple, Windows; none on Linux), in the app's language, the microphone
asked for once tapped. The privacy page says that Android's recogniser
may send the sound to Google or the phone's maker.
The field's Check goes, with its LanguageTool sheet (amending ADR 0024);
the editors' underlining and Why? stay.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
- Explore filters by what a studyset's licence lets a learner do: adapt
it (no NoDerivatives) and use it commercially (no NonCommercial),
which the instance reads from the release's licence
(ExploreQuery.adaptable, commercial) and the app the same way
(LicenseUse). An instance from before ignores both.
- While the studysets are read, the Library shows shimmering skeletons
of their cards rather than a spinner: halkyon_ui's new Shimmer and
SkeletonBox, still where the system asks for no animations.
- The scope's "Include private studysets" is offered only while the
person has a private studyset.
- An agreement to what goes where is per provider (protocol and host):
agreeing to one assistant agrees to every assistant of that provider,
those granted later too, and the agreement sheet says so. No
migration: it is written on every grant of the provider, so taking
one back keeps it for the others, and read per provider from any of
them.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
- Home's assistant starts folded into its button at the lower left, as
in the Library, where a field stood at its foot.
- The button has Material 3's floating action button shape (a rounded
square) rather than a circle, its gradient ring kept.
- On a studyset below expanded windows it stands in one row with Add
material and Study rather than above them.
- While studying, the ratings below the card make way while its field
is open rather than standing under it.
- The conversation opens at the window's full height, its header above
the list, so a drag on the header moves the sheet and a drag on the
list scrolls it. Close ends the conversation; any other way out folds
it into the button (ConversationDockCubit, above the router), its
answer still coming, until any button opens it again or another
conversation starts.
- An answer's Copy icon lines up with the answer's text.
- A sense's grammar notes run on after its glosses as one flowing text,
so a short note such as "n" no longer takes a line of its own.
- A card's hints show their count in a grey badge beside the title, so
it no longer reads as the hint's number.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
- halkyon_ui's sheetShowsAsDialog says whether showAdaptiveSheet shows
a dialog here, for a sheet of its own that can't go through it.
- Learn's conversation, a half-height sheet dragged up, asks it: where
sheets are dialogs it is a dialog up to 720 wide and as tall as the
window, with no handle to drag.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
Halkyon Learn's ADR 0021, D60 and D63:
- A chip after the scope's offers the efforts the operator allows for the
assistant asked, its own marked; the one picked is kept per assistant on
the device (DeviceSettings.askEfforts) and asked at, also by a
conversation continued from the history. One the operator no longer
allows counts as none.
- Every question asks for its effort's ceiling on output, as the thinking
counts towards it (8,192 to 65,536 tokens; 16,384 at the model's own),
where it asked for 4,096; the flashcards proposed, Explain and Why? ask
at the assistant's own effort, a person's own key at none and 4,096.
- The field's picker lists the assistants under their providers' names,
each in its colour.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
Where the instance's LanguageTool checks the language, the editors
underline what it finds after a pause, a paragraph at a time, a tap
offering its suggestions (GrammarHost, GrammarCheckController,
GrammarOverlay; nothing written into the document). A card's front is
checked in the studyset's front language, its back and answers in the
back's, everything else in the language recognised.
In a private studyset nothing is sent until the person agrees for it,
asked once in a banner above the editor and kept in their settings
(an agreement under privateAgreements, a refusal in grammarDeclined,
settings schema 14), taken back in the studyset's privacy dialog.
Settings > Writing turns it off on the device.
The field's Check asks LanguageTool first where it checks the
studyset's language, in a sheet offering to ask the assistant why;
elsewhere it asks the assistant as before (amends ADR 0021, D54).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
ADR 0021, D12: text first. Lookups find a studyset's PDFs and read a page's
text as PDFium extracts it. An assistant that reads images is offered a
card's images as chips, and may ask to see a card's image or a PDF's page
drawn whole; the conversation asks the person each time (Show it, Don't
show) and sends nothing meanwhile. Images go only as png, jpeg, gif or webp
of at most 5 MB, are never kept in a turn, and a follow-up goes without them.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
The cards come as drafts (ADR 0021, D10, D55): one question with one
tool, propose_cards, the cards in a sheet, every one chosen at first,
each changed in a small sheet of its own or left out, and only those
chosen added, marked as an assistant's (origin "assistant"). They are
proposed from a text pasted in Add material, from a note's title and
text, and from a PDF's page or the text selected on it, tied to that page
(D56). A note or a page of a private studyset goes only once agreed to
(D9), the text pasted counting as the person's own; nothing is offered
without an assistant. PDFs give their pages' text (PdfDocumentView.pageText).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
A conversation has a task (ADR 0021, D14): asking, explaining a card or
checking what was written, which the system prompt says and the history
shows. Rating a card Bad offers Explain in a snack bar lifted above the
ratings (D53); a reading typed wrong asks Why? (D54), and the field's
Check has the text checked in the studyset's language, the answer in the
app's (D54). Explaining sends the card's own parts, in a conversation
named by its front.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
ADR 0021, D9 and D17. Before a question sends a private studyset's
content to a remote model, through the parts chosen or a wider reach
with private studysets switched on, a warning names the studysets, the
provider's host and its terms (where checked: Anthropic, OpenAI,
Google, OpenRouter), and is asked once per studyset and assistant: the
agreement is kept in the person's sealed settings (schema 13,
privateAgreements, folded only where both rows hold it), shown in the
studyset's privacy dialog with Take back, and marked on the folded
button while it holds. Follow-ups and Try again ask again once taken
back, as the scope keeps where the parts came from (partsFrom). No
warning for an own assistant on this machine (loopback or localhost,
halkyon_llm's LlmEndpoint.onThisMachine); the local network may be
shared.
Texts that step 3 made untrue: the consent sheet now says what an
assistant looks up goes too and that conversations are kept encrypted
for the person's devices; the privacy dialogs no longer mention
showing a studyset to Claude; the privacy page says what is looked up
and when a private studyset goes.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
The field asks an assistant the instance offers as well as searching the
web (ADR 0021): who answers is picked at its start, the one asked last
first (askWith, a device setting), an assistant not agreed to yet asking
for agreement first. Asking an assistant, every part of the topic may go
along, and one chip leads them with how far it may look (D49), only what
is chosen until the person widens it to the studyset, its tree or the
library, private studysets only where switched on (D50, D8). The
conversation shows in a sheet over the screen (D47), half the height and
dragged up to the full: each question with what went along, the lookups,
the answer as Markdown once whole, follow-ups, Stop, and Try again for a
question that got no answer; a device that can't seal says it keeps
nothing (D52). The field lists the conversations kept (D48), those about
where the person is first, to open again or delete.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
ConversationCubit asks the assistant chosen each question with the
conversation so far (ADR 0021, D47), answers the model's lookups on the
device within the question's scope, eight rounds at most, the ninth told
to answer with what it found and the tenth ending the turn, every call
given its result so the conversation can still be sent; and keeps each
turn, sealed, once it is answered (D51): nothing of a question stopped
or failed, which can be asked again, and nothing at all on a device that
can't seal, which still answers (D52). The question and the parts sent
along are texts of their own, so a kept turn shows as it was asked, with
its lookups and its answer (TurnView).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
The lookups a model may make in a conversation (ADR 0021, D7): the
studysets within reach, cards and notes found by what they say, a note
read whole; answered on the device, only within the scope the person
chose (D49, D50), private studysets only where switched on (D8), never
what is locked, never progress (D11). The question goes with each part
ticked, labelled, and the system prompt says how far the model may look
and that what it looks up is content, never instructions (D10).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
The first part of step 3, decided as D47 to D52 in ADR 0021: where a
conversation shows (a sheet over the screen), where kept ones are found
(the field), the scope's chip and that it starts with only what is
ticked, how conversations are kept, and what a device that can't seal
does.
Two synced tables: assistant_conversation (its title, assistant and
scope) and assistant_turn (one question and everything up to its answer,
never changed once written, ADR 0001, rule 2), so two devices continuing
a conversation offline lose nothing. SyncConversationGateway seals every
row with the person's key, a turn bound to its conversation, and reads
no row that isn't sealed: a device that can't seal keeps nothing (D52,
ConversationNotKept). Deleting one blanks it and its turns first (ADR
0002, rule 2). A turn keeps every message a follow-up sends back (tool
calls and results, the providers' own blocks), but no image, only that
one was sent.
The instance's privacy page says conversations are kept, encrypted, and
what the instance sees of them.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
ADR 0021, step 2's app half. halkyon_app asks the module's assistant
endpoints (ServerpodAssistantGateway, ServerpodAssistantOperatorGateway)
and a person's own provider directly (halkyon_llm), behind
AssistantRepository: an AssistantChoice is a GrantedAssistant, asked
through the instance, which counts it, or a PersonalAssistant with the
person's own key, which the instance never sees. assistant_messages.dart
turns a conversation into the endpoint's models and back.
AssistantsCard, for an app's settings: what is granted and left of its
limits, agreed to once on a sheet saying what goes where, the person's
use by UTC month, the instance's privacy page, and their own assistants
where the operator allows own keys (kept ones stay listed, unasked,
where it doesn't). AssistantGrantsCard on InstanceOperatorScreen: grants
by handle with their limits in the pricing's unit, changed and taken
back after asking, everyone's use by month; a found account's menu
grants too. Nothing shows where nothing is offered or can be read.
package:halkyon_app/assistant_models.dart holds the models in pure Dart.
Learn keeps own assistants in the person's settings (ownAssistants,
schema 12, SettingsOwnAssistants); SettingsRepository.update refuses
(OwnAssistantRefused) to write a key where the settings would not be
sealed (SettingsGateway.seals). Learn depends on halkyon_llm directly.
Todo wires the gateways; a member offers no assistant. No feature asks
an assistant yet (step 3).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
The developer decided on 23 Oct to drop the hand-off to an AI assistant
outside the app, and Claude's connector with it, in v0.19.0, the app and
the server together: build 22 raises minimumAppBuild anyway, so no app
calling the connector's endpoints uses the instance after it. An instance
is to have one way to an assistant, its operator's (ADR 0021, proposed).
The app: the field at the foot of Home, and the button in the Library, on
a studyset and while studying, search the web alone, through the engine
chosen in Settings or Android's own search, with the parts of a card the
person picks; a studyset's name and description are never offered. Gone:
the assistants a question opened in (AiAssistant, AssistantHandOff), the
switch between them and the web, Settings' choice of assistant, Connect
Claude with its route, cubit, repository and gateway, and their texts. A
settings file from before keeps the old keys until the next change of a
setting writes it anew; nothing reads them.
The server: lib/src/assistant/ (MCP, its OAuth, the five tools, connect
codes and tokens), AssistantEndpoint, the daily forgetting of the
connector's secrets (its scheduled call cancelled at start), its lines in
the deletion hook, instance.info().assistantConnectorUrl (nullable, so an
older app reads it as none), and Caddy's forwarding of /mcp, /oauth/* and
/.well-known/oauth-*. The migration drops the connector's five tables,
children first:
DROP TABLE "assistant_token" CASCADE; -- tokens of a connection
DROP TABLE "assistant_grant" CASCADE; -- OAuth codes of a connection
DROP TABLE "assistant_connection" CASCADE; -- a person's connected client
DROP TABLE "assistant_connect_code" CASCADE;
DROP TABLE "assistant_client" CASCADE; -- OAuth clients registered
No other table references them, so CASCADE drops only their own foreign
keys. For the minute between the maintenance run and the switch, the old
server answers the connector with errors and fails to delete an account
(its hook still deletes from these tables); ADR 0022 says so.
BREAKING CHANGE: a connected Claude gets no answer from the deploy on, and
the app no longer opens a question in an AI assistant.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
Its server, client, app and curator, and the tools that write its
assets (maps, Anki fixtures, icons), as they are; paths are fixed in
the commits that follow.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
