If the AI assistant encounters an unexpected error at startup, it now displays a clear, actionable message rather than raw server output or technical tracebacks. Diagnostic details are preserved in logs for the team to investigate, but users see a clean prompt to try again.
