Notes from last discussion 2/2025:
currently, we realized that if we create a n obj inside built-in function, we don't need to pretend it's unfamiliar (no need to validate again; 'trusted' func);

• switching this to eval model might be tough
• if so, objective would be: to transmit a lot more info; validated and resolved status of objects
• also currently contract with Eval; all we need to know about the obj is 'there'; all processing is done before getting to Orch
• this would be in a virtual Eval; we don't need to worry about separating
• another big motivator is to migrate Eval or Orch to new language; would need to be in separate processes not necessarily separate containers
• TLDR: not necessarily move these to Eval but just need solution to heed 'separation of concerns'

Status/current decision: this should be descoped and we are not worrying about this this quarter

Ok, @ecarg, as the owner of this initiative, could you describe it? This means moving it to the Engineering backlog (outside of this quarter) and putting it in "open" again. And remove it as a dependency for this quarter "Epic", what do you think?