code wiki / _hdl_build / nx_ux_study_gate.nx

nx_ux_study_gate.nx

buildroot/runtime/_hdl_build/nx_ux_study_gate.nx

6600 B92 linesdepth 3pulls 3 transitivereach 0 importersview sourcekind gate/prooftopic ux
docsdependenciesstructsconstsfunctions

about

nx_ux_study_gate.nx -- PROOF (re-runnable, adversary-controlled) that the NN/g usability scoring in nx_ux_study is HONEST and cannot be gamed into a fake win, PLUS it runs the first REAL study: the Nishi browser dogfood of https://nishifamily.com/experiential, graded from the actual scroll recording (nishi_browser_f0..7.ppm) captured this session. The neg-controls are the load-bearing part: if someone "fixes" the grader to reward feature-presence, T-STICKER and T-CATAS go green-when-they-should-be-red and this gate turns RED. license_tier: ORIGINAL expect_exit: 0

dependencies 1 imports · 0 importers

nx_ux_study.nx nx_ux_study_gate.nx

imports: nx_ux_study.nx

imported by: nobody (leaf or entry point)

call flow from main pre-order; caps 40 nodes / depth 6 declared; ↻ = already shown

main ux_w sys_write mk_study ux_set_task ux_task ux_set_finding ux_finding chk ux_w ↻ ux_print_grade ux_w ↻ ux_grade ux_any_task_fail ux_task ↻ ux_max_severity ux_finding ↻ ux_task_success_rate ux_task ↻ ux_n sys_mmap sys_write ↻

structs

none

consts

none

functions

11func mk_study(goal: *u8, artifact: *u8, surface: i64, quant: i64, moderated: i64, ntasks: i64, nfind: i64) -> *UxStudy
called by 1: main
25func chk(name: *u8, got: i64, want: i64) -> i64
called by 1: main calls 2: ux_wux_print_grade
32func main() -> i64