fix(voice): move setup into Voice settings - #170

🤖 Automated code review

REQUEST_CHANGES: the exact three-dot comparison introduces two blocking issues in the Voice setup flow: it can bypass the unsaved-agent consent guard, and returning can silently cancel a completed setup because AppShell evaluates a separate readiness snapshot. Two non-blocking localization/responsiveness issues are also present. Static review covered every changed user flow, accessibility, i18n, navigation and consent guards, async/error/never-completes/lifecycle/race behavior, test honesty, project/design-system rules, and duplicate overlap. Supplied GitHub evidence was inspected; all listed check runs completed successfully, while the combined commit status was pending, so required checks still independently govern merge readiness.

Deterministic publication result: 2 blocking and 2 non-blocking finding(s) publishable; 0 duplicate(s) suppressed.

🤖 Automated code review

COMMENT: the fresh full review found two non-blocking async consistency issues. The prior four automated findings are fixed in the current comparison and were not repeated. Final self-check covered every changed Voice setup and return flow, model and Siri settings interactions, accessibility, English/Spanish localization, responsive behavior, navigation and consent guards, async/error/never-completes/lifecycle/race behavior, test honesty, design-system/project rules, duplicate overlap, and evidence plus user effect for severity. Supplied GitHub evidence was inspected and is structurally valid: DCO and clean-room install passed, while frontend and Tauri checks were still in progress and required checks independently govern merge readiness.

Deterministic publication result: 0 blocking and 2 non-blocking finding(s) publishable; 0 duplicate(s) suppressed.

🤖 Automated code review

APPROVE: the fresh full review found no publishable findings in the exact three-dot comparison. The six prior automated findings are fixed in the current code and were not repeated; all supplied threads are resolved and include substantive human-account replies. Final self-check covered every changed Voice setup, return, model-management, Siri selection, preview, download, and responsive settings flow; accessibility; English and Spanish localization; navigation and consent guards; async, error, never-completes, lifecycle, cancellation, and race behavior; test honesty; design-system and project rules; duplicate overlap; and evidence plus user effect for blocking severity. Supplied GitHub evidence was inspected and is structurally valid: every listed check run completed successfully, while the combined commit status remained pending, so required checks still independently govern merge readiness.

Deterministic publication result: 0 blocking and 0 non-blocking finding(s) publishable; 0 duplicate(s) suppressed.

Pending checks: 1 check(s) are not complete.

This approval reflects the completed code review only; merge readiness remains governed by the repository's required checks.