Replace deprecated ClickableText with LinkAnnotation.Url + Text (Compose
native URI handling). Add GET /api/v1/meta call after login and on each
app resume to detect API version skew — shows dismissible amber banner.
Audio-only user messages (no text, only audio attachments) now render as
a clean "🎤 Sprachnachricht" pill instead of the ugly filename chip
(voice_1782170416740.m4a). Detects via attachments.all { kind == "audio" }
&& content.isBlank(). Regular file attachments are unaffected.
When search mode is active and a vertex: model is selected, a second pill
appears next to "Suche" with a dropdown to choose between Google Search
(default) and Enterprise Web Search grounding. The selection is sent as
webSearchProvider in the chat request body. Matches the web frontend's
OptionPill for webSearchProvider.
Only visible for Vertex models — OpenRouter uses agentive web_search
server-tool which has no provider toggle.
Design: lower tintAlpha (0.80/0.86) for more blob bleed-through, accent-tinted
gradient borders + level1 shadow on active items, specular borders on search bar
and user card, accent gradient avatar, separator line in context menu.
Features: live search filter, "Titel generieren" (Sparkles, manual POST with
{ manual: true }), "Duplizieren" (copies all messages with new IDs + parentId
chain), titleGenerated field on ConversationSummary/Entity (Room v4).
Audio recording:
- Mic button wired to MediaRecorder (AAC/M4A, 44.1kHz/128kbps)
- Tap to start, tap again to stop — auto-uploads as audio attachment
- RECORD_AUDIO permission with runtime request
- Recording state: red stop-square on mic button
TTS read-aloud:
- Volume button on assistant messages → POST /api/v1/audio-gen
- Gemini-TTS generates speech server-side, MediaPlayer streams it
- Tap again to stop playback, accent-colored icon when playing
- Graceful error handling (Toast on failure)
Supporting changes:
- KaizenApi.generateSpeech() for /api/v1/audio-gen endpoint
- KaizenIcons.Volume2 added (Lucide)
- DE/EN strings for recording, permission, TTS states
- EmptyHero date respects system locale instead of hardcoded German
- Stop button (red square) replaces send/call during streaming, cancels stream cleanly
- Auto-logout on HTTP 401 (expired token) routes to login instead of error banner
- LoginScreen migrated from Modifier.shadow to kaizenShadow (design system pillar 4)
Security section in Settings with three cards:
- Biometric toggle (persisted in SecureStore, respected by unlock flow)
- Signed-in devices list with remote revoke (GET/DELETE /api/v1/auth/tokens)
- Password change form (POST /api/v1/auth/change-password)
requestBiometricOrPassword() now checks the biometric toggle — when
disabled, always falls back to password dialog.
Three paths in openConversation() called unlockAndOpen() without any
authentication when biometrics were unavailable. Sidebar ToggleLock
also sent PATCH { locked: false } without auth.
- Unified requestBiometricOrPassword() gates all unlock paths
- PasswordUnlockDialog as fallback when biometrics unavailable
- KaizenApi.requestUnlock() sends password to server for verification
- All unlock flows (open chat + sidebar toggle) now require auth
Reasoning pill (Brain icon + dropdown): 8 presets matching web
(Sofort/Aus/Minimal/Niedrig/Standard/Mittel/Hoch/Maximal).
"Standard" = no override, "Sofort" = no reasoning + throughput.
Sends reasoningEffort + preferThroughput in ChatRequest.
Sampling pill (Tune icon + popover): per-parameter toggle +
slider for Temperature (0-2), Top-P (0-1), Top-K (1-100).
Each parameter independently on/off — only enabled ones are
sent. Accent highlight when any param is enabled.
Both controls are separate from mode pills (Bild/Suche/Video)
and work in chat + search modes, matching the web layout.
ChatRequest extended with temperature/topP/topK fields.
Old ChatSpeed enum no longer used for reasoning (kept for
backward compat with ServerConfig).
Hardware-enforced via Android KeyStore + CryptoObject:
- AES key in TEE/StrongBox with setUserAuthenticationRequired(true)
- BiometricPrompt receives Cipher as CryptoObject, unlocked only
after real biometric auth (fingerprint/face), not bypassable
- Key auto-invalidated on biometric enrollment change
Flow: tap locked chat → BiometricPrompt → on success request
server unlock token (POST /api/v1/auth/unlock) → fetch messages
with X-Unlock-Token header → display conversation.
Dependencies: androidx.biometric:biometric:1.4.0-alpha02
Replaced the generic file picker with a dropdown menu offering three
options: take photo (camera preview), pick from gallery (image/*), or
choose any file (*/*). Camera captures are compressed as JPEG and
uploaded like any other file. Upload logic extracted into shared
uploadPickedFile helper.
Parse tool events into thinking/analyzing/inProgress/completed states
instead of showing all as generic "Tool" rows. Handles type "thinking"
(shows "Verarbeitet…"), "analyzing" (shows filename), "start"/"end"/
"error" with proper icons and elapsed time. Empty/thinking-only events
no longer show a confusing wrench icon.
StreamConsumer.Incremental now returns StreamState (content, reasoning,
tools, query, sources) instead of just visible text. All sentinel sections
are parsed and structured:
- \x01 REASONING: captured as separate string
- \x02 SOURCES: parsed as List<SearchSource> (title, url, snippet)
- \x03 TOOL: parsed as List<ToolEvent> (type, name, args, result, elapsed)
- \x04 QUERY: captured as search query string
UI components:
- ReasoningBlock: collapsible "Gedankengang" with expand/collapse animation
- ToolEventsBlock: shows tool-call steps with status icons (start/end/error)
- SourcesBlock: web search sources as compact cards with domain extraction
Message model extended with reasoning, tools, query, sources fields.
KaizenApi.chat() returns Flow<StreamState> instead of Flow<String>.
ChatScreen wires all fields to Message during streaming.
Fixed: empty append() on Incremental no longer prematurely ends tool phase.
StreamConsumerTest rewritten to verify all parsed sections.
Settings now persist across app restarts via SecureStore:
- Theme (KaizenThemeId), appearance (Light/Dark/System), username, locale
- All loaded on SettingsViewModel init, saved on every change
Server sync: fetchMe() returns name + email from /api/v1/me,
populates SettingsViewModel on login and each foreground resume.
Language selector: System / Deutsch / English pill toggle in Settings.
Locale preference stored; runtime switching deferred to next iteration.
MeResponse expanded to parse name + email (was defaultModel only).
SettingsViewModel tests updated for store-less default state.
German (default) in values/strings.xml, English in values-en/strings.xml.
Matches web app's next-intl de/en setup.
Migrated: greeting, hero headline, chat input, error messages, mode pills,
suggestion pills, sidebar (brand, search, dates, logout, settings),
model sheet, login screen (welcome, fields, errors, submit button),
conversation labels (locked, pinned).
ChatModels: Suggestion/ChatMode now use resource IDs instead of hardcoded strings.
greeting() is now @Composable (uses stringResource).
ModePillsRow: selected state uses resource ID (Int) instead of String.
Date labels (HEUTE/GESTERN/FRÜHER) localized via DateLabels data class.
