@phroi small update on the iCKB / CellScript benchmark we discussed earlier.
I have basically finished the first iCKB equivalence experiment in CellScript:
Thanks again for tightening the iCKB artefacts and review notes earlier. They were very useful when turning this from an idea into actual fixtures.
The current evidence records matching accept/reject behaviour, hashes, exit codes, cycles, tx-size/capacity/fee data, and named failure modes.
When you have time I’d really appreciate your view on whether the claim boundary is fair, especially around which iCKB branches should be counted and whether any wording is too strong. 