panel-mode ...

[flat-linex]

I've recently used several AI models as a team of PhD students to carry out a project (BDPD - Be Different Play Differential). In the course of the development I exploited several times the two following 'workflows':

1. panel inter model discussions: convene several models (Claude, Deepseek, Mimo, each running in a different instance of Claude-Code/Mimo-Code) around one open question, let them discuss via the shared transcript on disk, and converge on a single deliverable. Used in the ideate stage, and for multi-model bug hunts;
2. iterative sw debugging and paper refinement: multi step SW debugging and research paper validation using a leader model and other supporting models, relying on escalation to me for the issues that could not be settled by the procedure.

Would it be possible/useful to have similar workflows in MiMo-Code?

Looking forward your comments ...