Improve voice dictation reliability

[AmethystLiang]

Summary

• Speech stop handling now settles on stopped, error, exit, or timeout instead of waiting indefinitely on worker termination, and tracks the in-flight stop so concurrent stop/teardown calls coordinate on the same worker rather than racing.
• Reject stale warm-worker reuse when the selected model is no longer ready, and harden the reuse fast-path against a stop that starts concurrently on the same warm worker.
• Resolve speech worker model files with path.join instead of hardcoded / separators (cross-platform).
• Keep the stopping flag exception-safe so a failed postMessage can't permanently drop dictation audio.

Checks

• pnpm typecheck — passing
• pnpm exec vitest run src/main/speech/stt-service.test.ts — 22 passing
• pnpm lint runs in CI (local oxlint binary predates .oxlintrc.json's rule set, so the pre-commit lint hook can't run locally)

Electron validation skipped by request.

Risk: medium — touches speech worker lifecycle in the main process.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro

Run ID: c63fc6bb-c3b2-46e3-ab02-b38058c6a4a6

📥 Commits

Reviewing files that changed from the base of the PR and between 2cbec04 and 24ef286.

📒 Files selected for processing (1)

• src/main/speech/stt-service.test.ts

💤 Files with no reviewable changes (1)

• src/main/speech/stt-service.test.ts

---

📝 Walkthrough

Walkthrough

SttService now coordinates native worker stop operations through shared in-flight promises, timeout handling, worker identity checks, and outcome-specific teardown. Worker reuse is prevented during active stopping and when model readiness changes. Cloud stopping state cleanup is scoped with finally. Tests cover worker exit, errors, concurrent stops, duplicate stopped events, timeout behavior, and invalid warm-worker reuse. Model, token, and vocabulary paths now use path.join.

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Description check	⚠️ Warning	The description covers summary and testing, but it omits required Screenshots, AI Review Report, Security Audit, and Notes sections.	Add the missing template sections, including Screenshots, AI Review Report, Security Audit, and Notes, with the requested cross-platform validation details.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title is concise and accurately reflects the main change: improving voice dictation reliability.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[coderabbitai]

Actionable comments posted: 1

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro

Run ID: a3537949-69b5-453d-a1d4-7843889fe344

📥 Commits

Reviewing files that changed from the base of the PR and between f238952 and 2cbec04.

📒 Files selected for processing (3)

• src/main/speech/stt-service.test.ts
• src/main/speech/stt-service.ts
• src/main/speech/stt-worker-model-config.ts