[johnmatthewtennant]

Summary

The Spokesperson could only be chosen when a call started, with --mode expert-spokesperson. Switching between a Spokesperson call and a direct agent call meant ending the call and starting another.

• The Berd Call menu bar has a Spokesperson checkmark. Turning it off restarts the running call's session as a direct agent call, and turning it on switches back. The call keeps its control port, stream, mute state and input-during-TTS policy, as with settings --restart.
• With the Spokesperson on, its voice and speed sit under the checkmark, and the agents' voice, speed and input-during-TTS setting get the next section.
• berd-call settings --spokesperson true|false does the same from the CLI, through a new spokesperson_mode control request.
• Agents keep the voice they have now: the new session restores their live speech settings, including an OpenAI voice and model, which startup options can't express. Options only the other mode takes, like --stt-url or --realtime-url, are left out. In a --stream live call the Spokesperson comes back with --allow-direct-agent-speech, so agents keep their own voice beside it. Otherwise their speed is left out, since it would become the Spokesperson's, which can't go above 1.5×.

Related issue

None found.

Testing

• turning_the_spokesperson_on_restarts_the_call_with_the_agents_current_voice: a direct call started with Samantha, whose live voice is now Aaron at 1.25×, restarts in Expert-Spokesperson mode with Aaron at 1.25× restored, without its --stt-url, and with --allow-direct-agent-speech for a live stream. Without a live stream, a call at 2× restarts without its speed, and the options pass startup validation. Before the fix, that 2× call kept its speed, and the replacement session failed to start.
• turning_the_spokesperson_off_and_on_keeps_the_options_both_modes_take: off drops the realtime URL and the Spokesperson's voice, and on drops the STT URL.
• replies_with_their_own_voice_put_the_spokespersons_beside_it: the Spokesperson's voice and speed share the checkmark's section, and the agents' get the next one.
• menu_reflects_live_call_state: the checkmark is on in an Expert-Spokesperson call and offers to turn it off, and is off in a conventional call and offers to turn it on.
• In a live call started from the Buzz plugin (berd-call start --stream live), with Siri en-US-F at 2×, ticking Spokesperson moved the call to --mode expert-spokesperson --allow-direct-agent-speech, with en-US-F at 2× for the agents and the Spokesperson at 1×. Unticking it moved the call back to conventional at en-US-F 2×. The call stayed up the whole time, and status showed each state.
• The menu previously ticked the checkmark only once the new session answered. menu_reflects_live_call_state now covers a switch in progress, which shows the chosen mode at once.

[morgmart]

🤖 Automated code review

Engineering found one blocking behavior regression in the new Spokesperson-mode restart path. Supplied GitHub evidence is structurally valid and all ten check runs completed successfully at the reviewed head SHA; no local execution was performed.

Deterministic publication result: 1 blocking and 0 non-blocking inline finding(s) publishable; 0 duplicate(s) suppressed; 0 blocking screenshot-evidence requirement(s) in this review body.

[johnmatthewtennant]

🤖 Combined into #397 with its commits kept, so the stack no longer needs restacking and review covers one diff. Open review findings carry over there. (AI-generated comment)