Manage local Whisper setup, model integrity and engine lifecycle #22

Outcome

Explicit local-model setup and owned engine start/status/stop integrate with Agentflow CLI without changing native-agent routing or requiring cloud keys.

Acceptance criteria

• Pinned runtime/model information and checksum-verified atomic model setup/import are provided. Interrupted/corrupt installs preserve the prior valid model.
• No implicit model downloads during transcription, app login or microphone startup.
• Missing runtime/model, readiness, port conflicts and crashes are actionable status states.
• Start is idempotent; stop affects only this app's owned engine; configured external engines and Murmur's engine remain untouched.
• Offline restart and transcription work after setup; paths with spaces are tested.
• Login startup never begins microphone capture or agent work.

Required TDD workflow

For each behavior: write and run a focused failing test; confirm the failure is for missing behavior; implement minimally; refactor while green. Record actual red/green commands and outcomes. Use deterministic fixture transports/processes for contracts plus a real pinned engine/model smoke test. No personal sessions or recordings as test fixtures.

Completion evidence

Record commit/PR, test commands, engine/model versions and checksums, Mac hardware, real-engine results, and any unverified manual-device gates. Do not mark unverified acceptance criteria complete.

Implementation evidence

Implemented in #25. Automated acceptance, real offline engine results, and reproducible corpus are recorded in docs/local-whisper-qa.md and the PR. Native physical-device acceptance remains in the separate QA issue.