- Google Meet: avoid treating repeated participant words as multiple assistant-overlap matches when suppressing realtime echo transcripts. Thanks @vincentkoc.

- QA/cache: require the full `CACHE-OK <suffix>` marker before live cache probes stop retrying, so suffix-only prose cannot hide a broken probe response. Thanks @vincentkoc.

- Slack/Matrix: avoid creating blank progress-draft messages when `streaming.progress.label=false` and progress tool lines are disabled. Thanks @vincentkoc.

- Slack/Discord: suppress standalone tool-progress chatter when partial preview streaming has `streaming.preview.toolProgress: false`, matching the documented quiet-preview behavior. Thanks @vincentkoc.

- QA/Matrix: keep the mock OpenAI tool-progress provider aligned with exact-marker Matrix prompts so the hardened live preview scenario still forces a deterministic read before final delivery. Thanks @vincentkoc.

- Google Meet: make realtime talk-back agent-driven by default with `realtime.strategy: "agent"`, keep the previous direct bidirectional model behavior available as `realtime.strategy: "bidi"`, route the Meet tab speaker output to `BlackHole 2ch` automatically for local Chrome realtime joins, coalesce nearby speech transcript fragments before consulting the agent, and avoid cutting off agent speech from server VAD or stale playback pipe errors.

- Google Meet: suppress queued assistant playback and assistant-like transcript echoes from the realtime input path, so the meeting does not hear the agent's own speech as a new user turn and loop or cut itself off.