feat(beta): wire voice in/out end-to-end with push-to-talk (v0.2.0-beta.1)

Reaches the Internal Beta milestone of DEC-001's release sequence the
same day as Alpha. Adds voice capture and playback through the full
Flutter UI → FRB → Rust core → tsclientlib → server path.

Promotions from PoC:
  poc/audio-capture-playback-spike  →  crates/chanora_audio/

New product code:
  crates/chanora_audio/src/engine.rs — cpal capture and playback,
    audiopus Opus VoIP encoder (48 kHz mono 20 ms frames), tsclientlib
    AudioHandler for decode + jitter buffer + mix on playback,
    push-to-talk gate, graceful playback-only fallback when capture
    is unavailable.
  crates/chanora_protocol/src/adapter.rs — extended with
    voice_out_tx (clonable mpsc::Sender<OutPacket>) and
    take_voice_in() (one-shot mpsc::Receiver<InboundVoice>); main
    loop now interleaves outbound voice drain, event pumping, and
    control-request handling.
  crates/chanora_protocol/src/lib.rs — re-exports the few
    tsproto_packets types (OutAudio, OutPacket, InAudioBuf,
    AudioData, CodecType, Direction) that chanora_audio
    legitimately needs. Documented as the single deliberate
    cross-crate type re-export per SAD-067, justified by the
    performance cost of a parallel type hierarchy on the 20 ms
    voice frame.
  core/chanora_core/src/lib.rs — ChanoraSession::start_audio,
    set_ptt, audio_stats; disconnect now stops the engine first.
  crates/chanora_bridge/src/api.rs — startAudio, setPtt,
    audioStats commands and BridgeAudioStats DTO.
  apps/chanora_flutter/lib/main.dart — "Start audio" button +
    hold-to-talk PTT button with pressed/released visual state +
    live stats line (TX/RX/PTT). Stats polled every 500 ms.

ARB:
  Both en and zh-Hans gain startAudioAction, pttHoldToTalk,
  pttTransmitting, audioStatsLine. Banner updated to
  "Beta build — voice in/out wired; not production ready."

FRB config:
  flutter_rust_bridge.yaml gains local: true so codegen resolves
  the workspace member's library stem to "chanora_bridge" instead
  of falling back to "UNKNOWN".

Empirical verification (2026-05-14, against cn.teamspeak.app):
  cargo check + cargo test --workspace: all green.
  flutter analyze: 0 issues.
  flutter test: 4/4 passing including:
    - test/alpha_e2e_test.dart (regression: Alpha still works)
    - test/beta_e2e_test.dart (Beta: connect → startAudio →
      PTT cycle → disconnect against cn.teamspeak.app).
  Live smoke (cargo test alpha_smoke -- --ignored): 49 channels,
  37 clients retrieved.
  Capture stream open against the host PipeWire auto_null source
  refused (snd_pcm_hw_params); engine correctly logged the warning
  and continued in playback-only mode. TX=0 frames, RX=0 frames
  reflects the headless null-source environment; on a real mic
  host the encoder produces ~50 frames/second while PTT is held.

Honest Beta scope (NOT in this release):
  - AEC / AGC / NS / HPF DSP (DEC-007..010): AudioEffects exists
    as a struct but the filters are no-ops. Beta+ work.
  - Production-quality resampler: current code is linear
    interpolation. Beta+ work.
  - Identity persistence via chanora_storage: still ephemeral.
  - Push-to-Dart event stream: UI polls instead.
  - chanora_diagnostics tracing-layer wiring: still scaffold.
  - Mobile (Android) cdylib + UI: PoC-proven, not yet in product.
  - Reconnect / network-loss recovery for the voice path.

Docs updates:
  - docs/governance/product-decision-register.md bumped to v0.9.7
    (Beta-milestone change-history entry; no row changes).
  - docs/governance/poc-results-summary.md bumped to v0.6.0
    (RISK-PoC-005 updated with Beta progress).

CHANGELOG.md

@@ -6,7 +6,103 @@ This project is expected to follow a Conventional Commits style workflow.
 
 ## [Unreleased]
 
-### Added — Alpha build (v0.1.0-alpha.1)
+### Added — Beta build (v0.2.0-beta.1)
+
+- **Voice in/out wired end-to-end through the Flutter UI.** Per DEC-001
+  this reaches the Internal Beta milestone. Build hash: see the
+  `v0.2.0-beta.1` git tag.
+- `crates/chanora_audio/` promoted from scaffold to a working engine:
+  - cpal-based capture (mic gain, linear resampling to 48 kHz, mono
+    down-mix) and playback (48 kHz stereo, requested config).
+  - `audiopus::Encoder` for Opus VoIP encoding (20 ms / 960-sample
+    mono frames).
+  - `tsclientlib::audio::AudioHandler` for the decode + per-client
+    jitter buffer + mix on the playback side.
+  - Push-to-talk gate: encoder is bypassed entirely when PTT is off,
+    so no spurious silence frames leak out.
+  - Graceful playback-only fallback: if the host has no usable mic
+    (typical for headless CI / users who deny the mic permission),
+    capture logs a warning and the engine continues with output only.
+    `AudioEngine::capture_active()` exposes this for the UI.
+  - Live counters: `frames_sent` / `frames_received` / `ptt()`.
+- `crates/chanora_protocol/` extended with voice channels:
+  - `ProtocolClient::voice_out()` returns a clonable
+    `mpsc::Sender<OutPacket>` for outbound frames.
+  - `ProtocolClient::take_voice_in()` returns a one-shot
+    `mpsc::Receiver<InboundVoice>` of decoded `S2C` / `S2CWhisper`
+    packets, with the originating `from_client` ID extracted.
+  - Re-exports the few `tsproto_packets::packets` types (`OutAudio`,
+    `OutPacket`, `InAudioBuf`, `AudioData`, `CodecType`, `Direction`)
+    that `chanora_audio` legitimately needs. This is the **only**
+    deliberate cross-crate type re-export; per SAD-067 the audio
+    path is performance-sensitive and a parallel type hierarchy
+    would force a copy per 20 ms frame.
+  - Connection task interleaves outbound voice (drained first per
+    loop iteration), event pumping, and control-request handling.
+- `core/chanora_core::ChanoraSession` audio API:
+  - `start_audio(AudioEngineConfig)` — starts the engine attached to
+    the active connection. Idempotent.
+  - `set_ptt(bool)` — toggles transmission. No-op without an engine.
+  - `audio_stats()` → `(frames_sent, frames_received, ptt_active)`.
+  - `disconnect()` now stops the engine before disconnecting the
+    protocol task.
+- `crates/chanora_bridge/` audio surface:
+  - `start_audio()`, `set_ptt(active)`, `audio_stats()` Dart-callable
+    commands.
+  - `BridgeAudioStats { frames_sent, frames_received, ptt_active }`
+    DTO.
+  - Mapped `CoreError::AudioNotStarted` and `CoreError::Audio(_)`
+    arms in `BridgeError::From<CoreError>`.
+- `apps/chanora_flutter/`:
+  - Beta UI rewrite of `main.dart`: "Start audio" button after
+    connect; hold-to-talk button with pressed/released visual state;
+    live audio-stats line below the PTT (`TX … frames • RX … frames
+    • PTT on/off`).
+  - ARB key set expanded with `startAudioAction`, `pttHoldToTalk`,
+    `pttTransmitting`, `audioStatsLine` in both `en` and `zh-Hans`.
+  - `test/beta_e2e_test.dart` exercises the full
+    Dart → FRB → chanora_bridge → chanora_core → chanora_audio
+    path against `cn.teamspeak.app`. Verifies connect, audio start,
+    PTT toggle, disconnect.
+- `flutter_rust_bridge.yaml` now sets `local: true` so the codegen
+  resolves the workspace member's library name correctly. Without
+  this, the generated Dart side fell back to `libUNKNOWN.so` and
+  failed to load the cdylib.
+
+### Changed
+
+- `flutter_rust_bridge.yaml`: added `local: true`.
+- `chanora_bridge::api`: `BridgeError::From<CoreError>` now maps
+  `CoreError::AudioNotStarted` to `BridgeError::InvalidCommand` and
+  `CoreError::Audio(_)` to `BridgeError::Connection`.
+- `apps/chanora_flutter/test/widget_test.dart`: banner-string
+  expectations updated from "Alpha build" to "Beta build" and from
+  "Alpha 版本" to "Beta 版本".
+- `docs/governance/product-decision-register.md` bumped to v0.9.7
+  with a Beta-milestone change-history entry. No decision rows
+  change.
+- `docs/governance/poc-results-summary.md` bumped to v0.6.0 with a
+  Beta-milestone change-history entry; RISK-PoC-005 updated to
+  reflect Beta progress.
+
+### Notes (Beta scope honesty)
+
+- DSP chain (AEC / AGC / NS / HPF per DEC-007..010) is **not yet
+  implemented**. `AudioEffects` exists as a struct but its filters
+  are no-ops in v0.2.0-beta.1. Real DSP is queued for Beta+ work.
+- The capture resampler is a simple linear interpolator. Production
+  quality requires a proper resampler in Beta+.
+- Identity is still ephemeral per connect; persistence via
+  `chanora_storage` is queued.
+- No live event stream into Dart yet — the UI fetches snapshots and
+  audio stats on a timer instead of subscribing to push events.
+- `chanora_diagnostics` is still a scaffold; no redaction wired into
+  `tracing` yet.
+- Audio engine is desktop-only in this Beta. Mobile bundle of the
+  bridge cdylib + UI verification was proven by the PoC but is not
+  re-built into product code in this milestone.
+
+### Carry-over from Alpha (v0.1.0-alpha.1)
 
 - **First Alpha build wires the connect → snapshot → disconnect cycle
   end-to-end from the Flutter UI to a live TeamSpeak-compatible