What cost the most. I built before testing the brief’s claim that the seed in keeps.rs was the only thing keeping data parameters. The full cycle (build, then lower-std) came back empty, and only then did I start tracing. Stuck on the empty result, I went through nikaia run, then nikaia lower plus the nikaia.contracts file, then grepped the NK1102 text into check_arguments.nika, then lends.nika and emit_core_args.nika. That was about 8 turns of tracing that one probe would have saved.
lends_with and emit_arg_only_read never lend a data parameter of any method that takes a receiver;FnContract has no flag saying a method belongs to a trait;nikaia lower F.nika writes nikaia.contracts; nikaia run writes nothing I could inspect).cargo build and lower-std of the seed removal, the confused re-check that it really had built, and the rebuild to restore develop’s binary. That is three builds of about 40 s each for no information.it and tests/compiler.grep lends_in, takes_a_receiver). If the column you change does not reach that code, stop. The brief itself should have carried one verified probe line showing the claimed cause, for example: an inherent method with text: String already lowers to &str.