Introduce optional Apple account-backed credits with scoped gateway access while preserving local and BYOK paths. Refresh assistant behavior, tests, privacy disclosures, docs, and the website for the 2.0 experience.
Rotate AI idle suggestions with optional remote packs, move history/dictionary onto self-sizing Home preview cards, harden clipboard capture/prompting, and simplify keyboard chrome by dropping most liquid-glass shadows.
Add the AI keyboard tab, Agent settings, and user-owned LLM key path for 1.7.0, including streaming answers and web-search transports without the built-in DeepSeek fallback.
Keep marketing version at 1.6.6 and bump build to 59. Strengthen never-answer
polish safeguards, clipboard reply-intent continuity, voice undo UX, device
UITests harness, eval fixtures, and What's New assets.
Custom polish styles can opt in to emotion-matched emoji (default off), with
prompt-level opt-in detection so paste-only styles keep model-added emoji.
Also include Volcengine API-Key ASR auth, voice-processing capture, PiP flash
fix, and related keyboard Shift/haptics reliability work.
Use redacted cursor neighborhood and pause-aware chunks for more natural polish,
validate protected terms with retry/local fallback, and structure bilingual prompts
for consistency and provider prefix caching.
Unify Bailian/Volcengine/OpenAI realtime streaming, ABE polish routing with
fun styles, and a shared card-page Settings hierarchy; bump to 1.1 (build 32).
Fix cold-start overlay recursion that overflowed the main-thread stack when
recording began while the ready overlay was still up; also remove temporary
on-screen Flow DEBUG panels after the orange-mic investigation, and land the
macOS overlay/catalog/layout polish plus related Flow recovery hardening.
Reduce perceived latency from key release to final text:
- Adaptive chunking: 2.5s first chunk + 5s follow-ups so short
utterances start on-device recognition while still recording.
- Session-level ASR warmup and audio-format cache reuse to remove
per-utterance cold-start of SpeechAnalyzer.
- Mirror live pipelined partials to the keyboard transcript line via
a new flow.transcriptionPartial App Group key + Darwin ping.
Also commits the accumulated custom language model, Flow session,
keyboard extension restructure, and Xiaomi MiMo provider work in
progress on this branch.
Batch of in-progress app work from the working tree.
- feat(keyboard): CursorNavigation + CursorDragPad for caret movement;
KeyboardSoundFeedback for system key click sounds
- feat(dictionary): DictionaryAliasGenerator + PersonalDictionaryEntrySheet;
TranscriptPostProcessor quality gate; retire DictionaryLearner
- feat(ui): TabBarVisibility handling; drop PageHeaderRow /
PageHeaderConfirmButton; refresh views and localizable strings
- fix(security): move the hardcoded DeepSeek key out of
PreconfiguredKeys.swift into a gitignored PreconfiguredKeys.local.swift
(seeded from .example by generate-xcodeproj.sh)
- docs(agents): add Conventional Commits versioning + bilingual changelog rules
- chore(gitignore): ignore PreconfiguredKeys.local.swift, .cache/, pycache
Custom language model / lexicon work stays on
feature/custom-language-model-asr. Changelog bullets added under
[Unreleased]; no version bump.
v0.3.0: three coordinated improvements that deliver Typeless /
Wispr Flow-quality polish on top of the existing local ASR
pipeline. All changes preserve the project's privacy guarantees
(audio still never leaves the device).
## 1. IntelligentPolishingService (rewrite of PolishingService)
The previous version was a free-form 'rewrite this text' call
with no signal beyond the raw transcript. The new one is a
single LLM call that does three things in one pass, exactly as
Typeless and Wispr Flow do internally:
1. ASR error correction (homophones, near-misses, missing chars)
2. Polish (drop filler words, fix grammar, add punctuation)
3. Style adaptation per app context (code / email / chat / doc)
The merged-prompt design halves the round-trip vs the previously
proposed two-stage design (correction + polish separately) and
the academic literature confirms it performs equivalently for
everyday Chinese / English dictation.
## 2. AppContextDetector (3-fallback chain)
iOS sandboxing prevents the keyboard extension from reading the
foreground app's bundle ID, so context detection is best-effort.
The detector runs three fallbacks in order, with caching to
avoid the cold-start 'unknown' that would force a neutral-tone
LLM call every time the user opens a new field:
1. Heuristic on the text at the cursor (code / email / chat / doc)
2. 30-minute cache of the last successful detection
3. Time-of-day + weekend heuristic as a soft default
The keyboard extension runs the detector on every press of the
mic and persists the result to the App Group so the host app's
polisher picks it up.
## 3. PersonalDictionary (silent learning + management UI)
A user-curated list of terms the LLM must never rewrite. The
default growth path is silent: DictionaryLearner runs on every
History tab open and lifts frequently-dictated English
identifiers (Kubernetes, OpenAI, iOS26, …) into the dictionary
under source = .history. Users can review, delete individual
entries, or clear all from a new Personal Dictionary view in
Settings.
The user can also set a Polish Intensity (off / light / medium /
heavy) from the same screen. Default is medium, which is what
Typeless and Wispr Flow also use.
## Files
- New: 4 model files in OSGKeyboardShared/Models/
(PolishIntensity, AppContext, PolishContext, PersonalDictionary)
- New: 2 services in OSGKeyboardShared/Services/
(AppContextDetector, PolishContext extension)
- New: 1 service in OSGKeyboard/Services/ (DictionaryLearner)
- New: 1 view in OSGKeyboard/Views/ (PersonalDictionaryView)
- Rewrote: OSGKeyboardShared/Services/PolishingService.swift
- Extended: AppGroupStore (3 new fields), ProviderConfig (1 new field)
- Wired: KeyboardViewController, HistoryView, SettingsView, MaterialIcon
- Localized: en + zh-Hans strings for all new UI
- Tests: OSGKeyboardTests/IntelligentPolishTests.swift (16 tests)
## Verification
- All new code follows the existing Sendable / strict-concurrency
patterns (the keyboard extension stays within its 60MB sandbox;
the polisher remains an actor; @MainActor is applied to the
learner and the settings UI).
- Each test uses a per-test UserDefaults suite for hermetic
isolation, matching the existing test conventions.
- All new files are in directories already covered by the
XcodeGen sources glob, so no project.yml change is needed.
## Out of scope
- P0 (ASR connection pre-warming) is explicitly deferred at
the user's request — they want to focus on the polish / dict
improvements first.
- The Cloud polish (WebSocket) work is not touched.
## Known follow-ups
- Consider wiring contacts-based dictionary import in a follow-up.
- Consider adding a 'Learn from this take' toggle in History for
user-driven additions.
- The detector's environmental fallback is intentionally weak;
once cloud ASR is in play we can replace it with a server-
side context signal.
Rocky