Follow up on the mixed web/native Fleet code-health baseline adopted in #29. Current non-regression ceilings: web whole-source coverage 22.34% lines / 17.07% branches / 22.78% functions / 21.69% statements across 3,451 production lines; native production coverage 3,053/5,612 executable Swift lines (54.4013%) with 16 unit and 3 UI tests; 32 cross-language complexity violations across 1,185 functions and 11,637 function NLOC (max CCN 98, max function length 616, max 19 parameters); 18 exact clone groups / 234 duplicated lines across 81 web/native files; and one reviewed source suppression marker outside the generated agent payload. Prioritize hermetic coverage for large app pages, worker routing, local/demo persistence, and native presentation code before decomposing their largest functions. Lower each checked-in ratchet whenever the metric improves.
Follow up on the mixed web/native Fleet code-health baseline adopted in #29. Current non-regression ceilings: web whole-source coverage 22.34% lines / 17.07% branches / 22.78% functions / 21.69% statements across 3,451 production lines; native production coverage 3,053/5,612 executable Swift lines (54.4013%) with 16 unit and 3 UI tests; 32 cross-language complexity violations across 1,185 functions and 11,637 function NLOC (max CCN 98, max function length 616, max 19 parameters); 18 exact clone groups / 234 duplicated lines across 81 web/native files; and one reviewed source suppression marker outside the generated agent payload. Prioritize hermetic coverage for large app pages, worker routing, local/demo persistence, and native presentation code before decomposing their largest functions. Lower each checked-in ratchet whenever the metric improves.