Add demo trio personas and polish empty-state panel UX.

Harden Expert Persona builder and add model suggestion.

Build credentials during Phase 1 and streamline human participant setup.

perf(orchestrator): parallel phases, streaming, early table summaries

feat(conversation): pluggable structures + decision methods (Robert's Rules, ranked-choice, majority)

feat(resilience): multi-tier participant failure recovery + Response-priority toggle

feat: human participant turns + credential intake redesign

fix(frontend): default API_BASE to '' so Docker builds use same-origin

feat(transparency): add "View current chat prompts" to settings

feat(participants): add "Select N Automatically" option

feat(limits): expose repetition / failsafe limits in settings menu

Surface Credential Summary in settings + exports

Build CCAI Vibe Demo on top of LLMChats3 baseline

Import LLMChats3 master as CCAI baseline