code wiki / _hdl_build / nx_meet_trust.nx
nx_meet_trust.nx
buildroot/runtime/_hdl_build/nx_meet_trust.nx
about
nx_meet_trust.nx -- R6 of the NISHI MEET PLATFORM: the ANTI-FAKE engine (LIBRARY, no main). The headline
capability + the MEASURED exceed: rule-based detection of AI-generated / slop resumes and fake "ghost" job
postings -- the screening that lets "quality meet quality." Sovereign (nx_cc->nxasm, no ML runtime, no JS):
lowercase + weighted keyword signals (slop buzzword density vs concrete specifics: numbers, contacts, links,
salary, real URLs). Returns 0..100 (higher = more authentic); below TRUST_THRESHOLD => flagged. Thresholds are
named consts (rule 11). license_tier: ORIGINAL
dependencies 2 imports · 6 importers
imports: nx_syscalls.nxnx_meet_lib.nx
imported by: nx_meet_curate.nxnx_meet_curate_gate.nxnx_meet_signals.nxnx_meet_signals_gate.nxnx_meet_trust_gate.nxnx_meet_trust_run.nx
structs
| none |
consts
| 10 | const TRUST_THRESHOLD: i64 = 50 |
functions
| 13 | func trust_lc(src: *u8, n: i64, out: *u8) -> i64 |
| 25 | func trust_has(hay: *u8, n: i64, needle: *u8) -> i64 |
| 31 | func trust_has_digit(hay: *u8, n: i64) -> i64 |
| 38 | func trust_resume_score(text: *u8) -> i64 |
| 62 | func trust_is_fake_resume(text: *u8) -> i64 { if trust_resume_score(text) < TRUST_THRESHOLD { return 1 } return 0 } |
| 65 | func trust_posting_score(text: *u8) -> i64 |
| 88 | func trust_is_ghost_posting(text: *u8) -> i64 { if trust_posting_score(text) < TRUST_THRESHOLD { return 1 } return 0 } |