feat(voice): add native macOS Siri speech - #164
Conversation
morgmart
left a comment
There was a problem hiding this comment.
🤖 Automated code review
REQUEST_CHANGES: static review found one blocking lifecycle failure in the new Siri streaming path and one non-blocking window-cleanup issue. The supplied GitHub evidence was inspected; all reported check runs were completed successfully. No review threads were supplied, so no findings were suppressed as duplicates.
Deterministic publication result: 1 blocking and 1 non-blocking finding(s) publishable; 0 duplicate(s) suppressed.
morgmart
left a comment
There was a problem hiding this comment.
🤖 Automated code review
COMMENT: the fresh full-range static review found one new non-blocking settings-concurrency issue. Two still-present lifecycle issues are suppressed because they already have unresolved GitHub threads. The supplied GitHub evidence was inspected; all reported check runs were completed successfully.
Deterministic publication result: 0 blocking and 1 non-blocking finding(s) publishable; 2 duplicate(s) suppressed.
morgmart
left a comment
There was a problem hiding this comment.
🤖 Automated code review
NO PUBLICATION: the fresh full-range static review found no new publishable findings, but approval fails closed because a still-present settings-concurrency issue maps to an unresolved automation thread without a substantive human reply. A preview-specific manifestation of the Siri no-callback stall is also suppressed forever by the resolved thread's substantive human reply. The supplied GitHub evidence was inspected: the Linux Tauri Rust check failed, while other checks were still in progress; required checks independently govern merge readiness.
Deterministic publication result: 0 blocking and 0 non-blocking finding(s) publishable; 2 duplicate(s) suppressed.
morgmart
left a comment
There was a problem hiding this comment.
🤖 Automated code review
APPROVE: the fresh full-range static review found no new publishable findings. The previously reported stream-stall, window-ownership, and settings-concurrency issues are fixed in the current comparison. One remaining preview-specific no-callback manifestation is suppressed forever because it maps to a resolved thread with a substantive human-account reply. The supplied GitHub evidence was inspected: Linux, Windows, and macOS Tauri Rust checks, DCO, and clean-room install passed; frontend checks were still in progress, and required checks independently govern merge readiness.
Deterministic publication result: 0 blocking and 0 non-blocking finding(s) publishable; 1 duplicate(s) suppressed.
Pending checks: 2 check(s) are not complete.
This approval reflects the completed code review only; merge readiness remains governed by the repository's required checks.
Summary
Adds native macOS Siri voices as a speech-output backend. Pocket TTS remains the default. On macOS, users can explicitly choose Siri, and Berd automatically selects a downloaded catalog voice when needed.
Voice settings expose exact regional locales, system-provided previews before download, voice download and selection, playback speed, and refreshed installed state. Assistant text streams through the selected native voice using the existing voice-conversation lifecycle.
Reviewer-reproducible examples
Testing
Manually confirmed Siri assistant playback and playback-speed changes in the isolated macOS development build.